Skip to main content

Full text of "Year book - Carnegie Institution of Washington"

See other formats



Year Book 80 


Library of Congress Catalog Card Number 3-16716 

International Standard Book Number 0-87279-654-x 

Composition by Action Comp Co. Inc., Baltimore 

Printing and Binding by BookCrafters, Inc., Chelsea, Michigan 

December 1981 


Officers and Staff v 

Report of the President 1 

Reports of Departments and Special Studies 1 

Department of Plant Biology 3 

Department of Embryology 105 

Developmental Biology Research Group 229 

Geophysical Laboratory 247 

Department of Terrestrial Magnetism 427 

Mount Wilson and Las Campanas Observatories 579 

Program in Science Policy 651 

Administrative Reports 679 

Office of Administration 68 1 

Bibliography 683 

Report of the Executive Committee 685 

Abstract of Minutes of the Eighty-Fourth Meeting of the 

Board of Trustees 687 

Financial Statement 689 

Articles of Incorporation 705 

By-Laws of the Institution 709 

Index of Names 715 

Digitized by the Internet Archive 

in 2012 with funding from 

LYRASIS Members and Sloan Foundation 

President and Trustees 

James D. Ebert 

William R. Hewlett 

William C. Greenough 

William T. Golden 

Philip H. Abel son 
Robert O. Anderson 
Lewis M. Branscomb 
William T. Coleman, Jr. 
Edward E. David, Jr. 
John Diebold 
Gerald M. Edelman 
Carl J. Gilbert 
Robert G. Goelet 
Crawford H. Greenewalt 
Caryl P. Haskins 
Richard E. Heckert 
John D. Macomber 1 
William McChesney Martin, Jr. 
Antonia Johnson Morner 
Franklin D. Murphy 
Robert M. Pennoyer 
Richard S. Perkins 
Robert C. Seamans, Jr. 
Frank Stanton 
Charles H. Townes 
Juan T. Trippe 2 

Henry S. Morgan 
Garrison Norton 
Charles P. Taft 
James N. White 3 
Trustees Emeriti 

'Elected May 1, 1981. 

2 Died April 4, 1981. 

3 Died September 9, 1981. 

Former Presidents and Trustees 



Vannevar Bush 1939-1955 

Caryl P. Haskins 1956-1971 

Philip H. Abelson 1971-1978. 

Robert A. Lovett 1948-1971 

SethLow 1902-1916 

Wayne Mac Veagh 1902-1907 

Keith S. McHugh 1950-1974 

Andrew W. Mellon 1924-1937 

John Campbell Merriam 1921-1938 
Margaret Carnegie Miller 1955-1967 

Roswell Miller 1933-1955 

Darius O. Mills 1902-1909 

S. Weir Mitchell 1902-1914 

Andrew J. Montague 1907-1935 

William W. Morrow 1902-1929 

SeeleyG. Mudd 1940-1968 

William I. Myers 1948-1976 

William Church Osborn 1927-1934 

Walter H.Page 1971-1979 

James Parmelee 1917-1931 

Wm. Barclay Parsons 1907-1932 

Stewart Paton 1916-1942 

George W. Pepper 1914-1919 

John J. Pershing 1930-1943 

Henning W. Prentis, Jr. 1942-1959 

Henry S. Pritchett 1906-1936 

Gordon S. Rentschler 1946-1948 

David Rockefeller 1952-1956 

ElihuRoot 1902-1937 

Elihu Root, Jr. 1937-1967 

Julius Rosenwald 1 929- 1 93 1 

William M. Roth 1968-1979 

William W. Rubey 1962-1974 

Martin A. Ryerson 1908-1928 

Henry R. Shepley 1937-1962 

Theobald Smith 1 9 1 4- 1 934 

John C. Spooner 1902-1907 

William Benson Storey 1924-1939 

Richard P. Strong 1934-1948 

William H. Taft 1906-1915 

William S. Thayer 1929-1932 

JuanT.Trippe 1944-1981 

James W. Wadsworth 1932-1952 

Charles D. Walcott 1902-1927 

Frederic C. Walcott 1931-1948 

Henry P. Walcott 1910-1924 

Lewis H. Weed 1935-1952 

William H. Welch 1906-1934 

Andrew D. White 1902-1916 

Edward D. White 1902-1903 

Henry White 1913-1927 

James N.White 1956-1979 

George W. Wickersham 1 909- 1 936 

Robert E. Wilson 1953-1964 

Robert S. Woodward 1905-1924 

Carroll D. Wright 1902-1908 

Under the original charter, from the date of organization until April 28, 1904, the following 
were ex officio members of the Board of Trustees: the President of the United States, the Presi- 
dent of the Senate, the Speaker of the House of Representatives, the Secretary of the Smithso- 
nian Institution, and the President of the National Academy of Sciences. 

Daniel Coit Gilman 1902-1904 
Robert Simpson Woodward 1904-1920 
John Campbell Merriam 1921-1938 

Alexander Agassiz 1904-1905 
Lord Ashby of Brandon 1967-1974 

J. Paul Austin 1976-1978 

George J. Baldwin 1925-1927 

Thomas Barbour 1934-1946 

James F.Bell 1935-1961 

John S. Billings 1902-1913 

Robert Woods Bliss 1936-1962 

Amory H. Bradford 1959-1972 

Lindsay Bradford 1940-1958 

Omar N. Bradley 1948-1969 

Robert S. Brookings 1910-1929 

Vannevar Bush 1958-1971 

John L. Cadwalader 1903-1914 

William W. CampbeU 1929-1938 

John J. Carty 1916-1932 

Whitefoord R. Cole 1925-1934 

John T.Connor 1975-1980 

Frederic A. Delano 1927-1949 

Cleveland H. Dodge 1903-1923 

William E. Dodge 1902-1903 

Charles P. Fenner 1914-1924 

Michael Ference, Jr. 1968-1980 

Homer L. Ferguson 1927-1952 

Simon Flexner 1910-1914 

W. Cameron Forbes 1920-1955 

James Forrestal 1948-1949 

William N. Frew 1902-1915 

Lyman J. Gage 1902-1912 

Walter S. Gifford 1931-1966 

Cass Gilbert 1924-1934 

Frederick H. Gillett 1924-1935 

Daniel C. Gilman 1902-1908 

Hanna H. Gray 1974-1978 

Patrick E. Haggerty 1974-1975 

John Hay 1902-1905 

Barklie McKee Henry 1 949- 1 966 

Myron T. Herrick 1915-1929 

Abram S. Hewitt 1902-1903 

Henry L. Higginson 1902-1919 

Ethan A. Hitchcock 1902-1909 

Henry Hitchcock 1902 

Herbert Hoover 1920-1949 

William Wirt Howe 1903-1909 

Charles L. Hutchinson 1902-1904 

Walter A. Jessup 1938-1944 

Frank B. Jewett 1933-1949 

Samuel P. Langley 1904-1906 

Ernest O. Lawrence 1944-1958 

Charles A. Lindbergh 1934-1939 

William Lindsay 1902-1909 

Henry Cabot Lodge 1914-1924 

Alfred L. Loomis 1934-1973 



1530 P Street, N.W., Washington, D.C. 20005 

James D. Ebert President 

Robert N. Kreidler Vice President 

James W. Boise Bursar; Executive Secretary to the 
Finance Committee 

Christopher Wright Staff Member (Science Policy and 
Institutional Development) 

Ray Bowers Publications Officer; Editor 

Montgomery S. Bradley Special Assistant, Corporate Sponsors 


Kenneth R. Henard Assistant Bursar 

Susan Y. Vasquez Assistant to the President 

Joseph M. S. Haraburda Assistant to the Bursar 

Richard E. Hewitt Administrative Officer for Services 

Marshall Hornblower Counsel 


Roy J. Britten 


Barbara McClintock 


Carnegie Institution of Washington adheres in all phases 
of its operations, including employment and educational 
programs, to a policy barring discrimination on the basis 
of race, religion, color, national or ethnic origin, sex, or 
physical handicap. In its educational programs it admits 
qualified students as fellows without regard to race, reli- 
gion, color, national or ethnic origin, sex, or physical 
handicap to all the rights, privileges, programs, and ac- 
tivities generally accorded or made available to fellows at 
the Institution. It does not discriminate on the basis of 
race, religion, color, national or ethnic origin, sex, or 
physical handicap in administration of its educational 
policies, admissions policies, fellowship programs, and 
other Institution-administered programs. 

Report of 
the President 


50 um 

Often characterized as the ultimate biological challenge and of 
key interest to virtually all scientists is the way in which nerve 
cells are specifically linked together to form the neural net- 
works that underlie the phenomena of perception, memory, and 

James D. Watson 

The history of science has shown that a single improvement in 
instruments may render years of work with inferior apparatus 

George Ellery Hale 

Carnegie Institution of Washington Year Book 1 


■*■ ington, made in accordance with the By-Laws, recounts a very 
satisfying year on all fronts— a year rich in discovery and in striking 
out in new directions. Also noteworthy has been our continuing im- 
provement in matters financial: For the first time in more than a 
decade, the Institution's revenues last year exceeded expenditures. 

Our satisfaction over our financial improvement must be kept in per- 
spective. Our financial house is in order, but our resources are still too 
limited. Presidents of academic institutions can sometimes become too 
engrossed with the search for financial resources. I am no exception, 
for I know that stability of funding, no less than vision and daring, are 
crucial in science. Still, in our case there is little danger that the scien- 
tific imagination of the president will limit the creativity of our 
research, since our directors and staff members— our "living endow- 
ment"— continue to display boldness, curiosity, and ingenuity in abun- 
dance. Thus I shall continue to invest a large part of my energies 
toward ensuring the continuity of their support, enabling them to 
work, as far as possible, without diversion. 

There is wide concern that the weakness of the nation's economy 
may force reductions in the scientific enterprise. I do not fear a reduc- 
tion in overall size; perhaps we have been too long wedded to massive- 
ness. I do fear a partial moratorium in the creative output of basic 

(Opposite) Kenneth Muller and colleagues at the Department of Embryology are investigating 
nerve cell interactions and their capacities for repair after damage in the medicinal leech. Shown 
at left is a photomicrograph of nerve cord linking two nerve cell clusters (ganglia 9 and 10), 27 
days after axons were severed by crushing and their ensheathing cells killed. The lower axon is 
regenerating toward junction with the upper axon at about the midpoint. Electrical synapse is 
restored by such regeneration in about 80 percent of such cases. In this case, the regenerating 
axon (enlarged at right) has established a weak electrical connection with the upper axon via the 
remains of the severed axon thereby linking the ganglia even before axonal regeneration has been 
completed. (See page 42-43.) 


science, unless the scientific community can agree upon, and persuade 
the federal government to accept, ways of buttressing those lines of 
research where promise is greatest. The Carnegie Institution's historic 
willingness to make hard choices in order to strengthen new directions 
exemplifies this rule. Corollary guidelines are also to be found in 
Carnegie's record: We continue no program merely because it has been 
done in the past, and we allow no labels to constrain us unduly. 

Instrumentation for Tomorrow's Discovery 

There is another point where the Institution's and the nation's needs 
coincide: the need constantly to improve the instruments that are the 
essential tools of the scientist's trade. Instrument requirements vary 
greatly among and even within disciplines, but the crucial role of in- 
strumentation in astronomy is illustrated dramatically in George 
Preston's introduction to his Annual Report as director of the Mount 
Wilson and Las Campanas Observatories, and in his essay, "The role of 
instrumentation in astronomy," from which I shall draw heavily. 

Knowledge of the extraterrestrial universe is derived from the study 
of photons that reach the Earth in trickles from celestial sources. 
Telescopes are built to concentrate photons for analysis by a variety of 
instruments. These instruments, the equivalents of laboratory devices 
in other disciplines, play a central role in observational astronomy 
because they limit and to a great extent color the character of the 
research that can be undertaken. But they are costly. The confines of a 
finite research budget require thoughtful decisions about what instru- 
ments to build. Major considerations are the technical feasibility of the 
proposed device, and the scientific objectives and instrumental skills 
of the scientific user-group. Once constructed, many astronomical in- 
struments can be improved as new technology permits. 

Not all instrumental improvements are stimulated by new technol- 
ogy. Some apply long-known principles but in new ways, specifically to 
solve age-old problems. George Ellery Hale, founder of Mount Wilson 
Observatory, in 1902 observed, "The spectra of red stars have been 
found to resemble in a most interesting way the spectra of sunspots. 
On the strength of this, one would be tempted to conclude that these 
stars are covered by spots like those of the Sun ..." 

Thirteen years later, Hale summarized his discoveries at the Mount 
Wilson solar towers— his explorations of "solar meteorology" and his 
discovery of magnetic fields near sunspots. Reviewing stellar prob- 
lems, he returned to his earlier speculation: "Thousands of stars in the 
same stage of evolution doubtless exhibit similar phenomena which are 


hidden from us by distance .... The simplest and most obvious step 
would be to make just such a study of the physical phenomena of 
selected stars ... as we are now making of the Sun." 

Preston points out that these sentences of Hale, many years ago, are 
prophetic— "prologue to the remarkable discoveries made during the 
past year by Arthur Vaughan and his colleagues at Harvard- 
Smithsonian Center for Astrophysics, Sacramento Peak Observatory, 
and the University of Utrecht. . with the Mount Wilson 1.5 -meter 

In his pioneering studies of solar-like activity cycles in lower-main- 
sequence stars, Olin Wilson had detected "noise" in excess of his mea- 
surement errors and speculated that it might correctly be attributed to 
rotational modulation of observed chromospheric activity. Rotational 
modulation would occur if persistent irregularities existed in the 
longitudinal distribution of active regions on the surface of a star. Such 
irregularities have been found on the surface of the Sun by Howard 
and his colleagues at Mount Wilson. Thus, to an observer on Earth, 
chromospheric emissions should appear to increase and decrease as ac- 
tive regions rotate into and out of the field of view. But Wilson, who 
could observe only a few nights per month, saw only uninterpretable 

To study stellar activity cycles and rotations, Vaughan designed and 
constructed a special spectrometer to measure those line-emission 
fluxes of ionized calcium that are the signatures of magnetic activity 
on the surfaces of cool stars. Wilson, in his earlier work, with great ef- 
fort had set up his experiment monthly at a large multipurpose spec- 
trograph at the Mount Wilson 2.5-meter telescope. Vaughan 's device 
was relatively simple, was dedicated solely for use at the Mount Wilson 
1.5-meter telescope, and was capable of accuracy equal to that achieved 
by Wilson even when used by nonprofessional operators. Because ma- 
jor programs could now be undertaken by making routine what was 
previously complex, Vaughan 's group made several important discov- 
eries in the last two years. The instrument provides an excellent exam- 
ple of design with forethought and purpose. 

Image Intensifiers and Detectors. In contrast, technological im- 
provements—in detector sensitivity, for example— are often the stimu- 
lus for new instruments. Major improvements in the quantum effi- 
ciency of indium-antimony photocathodes in the 1-5 m region of the 
infrared induced Eric Persson, in collaboration with Tom Geballe, to 
construct a cold-grating Fabry-Perot spectrometer. This device will 
greatly aid the study of the activity that takes place in regions of star 
formation embedded in dark clouds. The clouds of particulate matter 
are opaque to visible light but become increasingly transparent with 


increasing wavelength. Thus the new infrared spectrometer will pene- 
trate the clouds to probe the excitation and kinematics of the hot in- 
terior gas and the presently unexplained sources of ionizing radiation. 

Techniques for analysis (photometry, spectroscopy, interferometry, 
polarimetry, and so forth) have been explored intensively during the 
past two decades and are now fairly well in hand. Emphasis through- 
out the 1980's will shift to new detectors that will record photon events 
after analysis. 

The first (non-human) detector in astronomy was the photographic 
emulsion. Because emulsions can be many centimeters in dimension, 
they will continue to be employed for wide-field imaging. But their low 
quantum efficiency— only a few percent of incoming photons can be 
recorded— and limited dynamic range restrict their usefulness for 
many applications. The photocathodes employed in photomultipliers 
and image intensifiers have efficiencies an order of magnitude larger, 
but these cannot be constructed to record large fields of view. There is 
now appearing a new generation of silicon chips with quantum efficien- 
cies that approach unity. These devices are now undergoing intensive 
scrutiny by astronomers. 

Ordinarily, the astronomy community cannot sustain the research 
and development effort of the magnitude required to develop such 
kinds of photovoltaic devices. (The Carnegie Image Tube was a notable 
exception.) It must rely on private industry, which provides related 
products largely for military and communications applications. Adapt- 
ing commercially developed devices for astronomical purposes can be 
difficult, but it is also rewarding. A few years ago, the linear Reticon 
array was such an innovation. By placing image intensifiers in se- 
quence in front of the Reticon, and providing key ideas to make the 
system a reliable photon counter, Stephen Shectman produced a highly 
successful detector system for astronomical spectroscopy at low light 
levels. The system, installed at the Mount Wilson 2.5-meter telescope, 
has been replicated with improvements at the du Pont 2.5-meter tele- 
scope and, most recently, at the 5-meter Hale telescope at Palomar. In- 
dicating the eagerness with which the community has embraced this 
innovation, the "Shectographs," as they are now called, were used on 
more than 400 nights at the above three telescopes during the past 
year, and the system has been reproduced by groups at the University 
of Michigan, the Harvard-Smithsonian Center for Astrophysics, the 
University of Hawaii, and the Mount Stromlo Observatory in Austra- 
lia. Clearly Shectman 's work with the Reticon has had a major impact 
on optical astronomy. 

Silicon chips are outstanding converters of photons to electric 
charge, but the total charge at each position in a faint astronomical im- 
age is so small— a few thousand electrons at most— that the problem of 


measuring the charge pattern in an image has been very difficult to 
solve. Interest at major observatories is now centered upon the appli- 
cation of CCD's— integrated circuit television cameras. The expression 
CCD stands for charge-coupled device, which refers to the method used 
for manipulating the charge pattern resulting after exposure to an op- 
tical image. If care is taken in the design of external clocking and 
amplifier circuits, the CCD structure can be used to measure the 
charge at each location in an image of typically 500 by 500 elements, 
with an accuracy of from 10 to 100 electrons, depending on the par- 
ticular device. 

Much of the excitement over CCD's stems from the prospect of using 
them for discriminating measurements over an extremely wide field. 
At Las Campanas, for example, Carnegie has two telescopes equipped 
for wide-field photographic imaging. The photographic plates are very 
large, up to 50 X 50 cm 2 , much larger than the CCD's, which are 
typically 1 cm 2 . However, the CCD is from 10 to 100 times more effi- 
cient in recording photons. Moreover, the CCD can be used over and 
over again, to make accurate calibrations of the response of the device 
at each location. The photometric accuracy of the CCD is better than 
one percent, an order-of-magnitude improvement over photographic 

When starlight is measured through broad-band filters, for example 
in the red, blue, and ultraviolet portions of the spectrum, the ratio of 
the fluxes in two bands (say red/blue) exhibits a range of a factor of 
three, from the coolest to the hottest stars. But for stars of a given 
temperature, the photometric effects in a third color band caused by 
differences in age or in chemical composition are at most a few percent. 
It is precisely the measurement of such small photometric effects that 
has led to the present limited understanding of the history of star for- 
mation and the synthesis of the chemical elements in our Galaxy and 
other galaxies. These measurements have required greater accuracy 
than can be achieved by photographic techniques, and have been made 
one star at a time using single-element photomultiplier tubes. Can a 
CCD be designed capable of simultaneously measuring an entire field 
of stars? 

It is just such a multiband imaging photometer that Shectman 
wishes to build for Las Campanas. He also hopes to incorporate, for the 
first time in any CCD, a capability for measuring the starlight in a field 
simultaneously in several colors. The excellent seeing and photometric 
transparency of the atmosphere at Las Campanas make it an ideal site 
for such an instrument. 

A dichronic beamsplitter can be used to separate blue and red light 
en route to the telescope focal plane. A pair of CCD's, one at the blue 
focus and one at the red focus, can be used to measure the same field 


simultaneously in two bands. This makes the flux ratio insensitive to 
small changes in atmospheric transparency. Since the focal plane of 
the Las Campanas telescopes is so large, several such simultaneous 
pairs can be set up to measure different wavelength bands. The tele- 
scope motion can be controlled so that each of the pairs is exposed to 
the same strip of sky in succession. 

The problem with any of the next generation of detectors is that the 
volume of information is so large that none of the data-recording 
systems available at Carnegie telescopes currently are adequate to the 
task. This will be remedied by the acquisition of fast minicomputers 
with large memories. A similar problem exists in analyzing direct 
photographs obtained with the du Pont and Swope reflectors, both of 
which were designed to have exceptionally wide fields of view. The 
1.5-degree field of the du Pont telescope is recorded on 50 X 50 cm 2 
plates. Under the excellent seeing conditions that commonly occur at 
Las Campanas, such a photograph contains approximately 10 8 re- 
solved picture elements, i.e., it would take 10 8 numbers to describe the 
entire photograph quantitatively. The potential of such photographs is 
immense for astrometry, colorimetry, variable star searches, star and 
galaxy counts, surface photometry of extended sources, and other 
uses. A scheme whereby photographs can be digitized for study with 
an appropriate computer is now in the planning stage at the Obser- 

The lesson of this brief account is clear: It is imperative that the 
scientific community have the means to provide instrumentation ade- 
quate to keep pace with the constant flow of new ideas. It is not clear, 
however, why we, as a nation, have ignored this truism in recent years, 
unless we are not sufficiently concerned about the integrity and 
strength of our scientific scholarship. 

The Integrity and Strength of Science 

Greatness in a nation, like personal greatness, is a measure not 
only of character, not only of excellence, but also of enduring 
significance both in ideals and in the shape and goals of effort. 
In any society, but perhaps especially in our own, such great- 
ness must necessarily be valued in coin of our own design and 
minting, whose very significance, moreover, we must ourselves 

.... And since a society can achieve and maintain greatness 
only if its people as a whole demand and recognize and judge 
greatness with an accurate and critical and uncompromising 
eye, it is worth a good deal of trouble to us as a nation to strive 


deeply to comprehend the nature of scientific scholarship in its 
truest sense. 

Caryl P. Haskins 

Carnegie Institution of Washington Year Book 62 

Are we as a nation sufficiently concerned about the integrity and 
strength of our scientific scholarship? 

I believe that the totality of the scientific enterprise— scientists and 
scientific institutions— must share in the reductions in national budget 
designed to restore a healthy economy. But I believe equally firmly 
that decisions about the programs where reductions are to be effected 
require full partnership of government and science. In his thoughtful 
essay in this volume, Christopher Wright speaks of Vannevar Bush's 
emphasis on the government's need to recognize the independent 
nature of the research community and the "need to fashion a legiti- 
mate bridge or compact, as among equals, between that dynamic sci- 
ence community and the federal government." Nowhere is the need for 
such a compact more crucial than in the setting of priorities. Yet, even 
the most casual observer of the budget-making process in recent 
months would have to conclude that the independence, disinterested- 
ness, and capacity for objective analysis that can be provided by the 
scientific community have been largely ignored. 

There are even more telling signs of this condition than in the failure 
to provide adequate support for scientific instrumentation. I believe 
that it was not a coincidence in fiscal year 1982, at a time of resurgence 
of the pseudoscience of "creationism," that the budget proposed for 
the National Science Foundation eliminated its Science Education 
Directorate. It was the Science Education Directorate that had sup- 
ported not only the biology textbooks developed by the Biological 
Sciences Curriculum Study, all of which treat evolution as a major 
biological theme, but also Man— A Course of Study, the social sciences 
curriculum which stressed the evolutionary origins of human nature. 

We have seen no more flagrant form of repression in this nation in 
many years than the moves toward eliminating the support of research 
in the behavioral and social sciences and toward restricting the 
teaching of evolution in our schools by requiring "equal time" for the 
teaching of a pseudoscience. This so-called "scientific creationism" 
charges that evolution theory is itself a religious statement, a form of 
secular humanism; it meanwhile argues that the universe is very 
young, that all animals and plants were created at one time, about ten 
thousand years ago, and that the Grand Canyon and indeed all fossils 
were created in a single year following the Worldwide Flood and were 
deposited in the order in which fossils are now found as the Flood rose 
and ebbed. Clearly this is not enlightened theology. 


Need we be alarmed? My answer is a simple but unequivocal yes. As 
Richard Lewontin wrote in BioScience, 

The teaching of the truth about the evolution of life on earth is under 
a more powerful and effective attack than at any time in this century. 
The creationist know-nothings, with only partial success in legislat- 
ing ignorance state-wide, have chosen the individual locality and 
school board for their political action. They are well aware that 
biologists are simply not numerous enough, nor sufficiently well in- 
formed of the local situation, to appear at individual school district, 
county and municipal hearings to explain the facts of evolution. 

Lewontin emphasizes, properly, that creationists have capitalized on 
scientific disputes among biologists on the details of the evolutionary 
process. He goes on to say, "Evolutionary study is a living science; as 
such it is rich with controversy about particular issues of detail and 

How should the science community respond? Wright has identified 
four distinguishable mediating functions between science and govern- 
ment: to manipulate scientific knowledge for practical ends; to estab- 
lish a place for the scientific enterprise, and to gain support for scien- 
tific research and for the involvement of science in government; to 
assess the means for achieving policy objectives; and to alter or create 
new goals or policies. None of these mediating functions precisely fits 
the role of science in combating ignorance and pseudoscience, unless it 
is the first. It is certainly as important to provide warnings about 
pseudoscience as it is to provide warnings against bad weather. It is 
not a new role for the scientific community, indeed for the Institution's 
own scientists. Wayne A. Moyer has pointed out that in 1938 Oscar 
Riddle, of the Institution's Department of Genetics, then first Presi- 
dent of the National Association of Biology Teachers, wrote in the 
premier issue of the American Biology Teacher, "... that a citizenship 
acquainted with the principles and subject matter of the life sciences is 
vital to our national welfare." But today it is not just the life sciences 
that are endangered, but the earth sciences and astronomy as well. To 
paraphrase Moyer, if all living creatures were created just ten thou- 
sand years ago, why study the age of the Earth, measured in billions of 
years, not thousands? Why seek further evidence of drifting con- 
tinents, with periods of mountain building and erosion spreading over 
long stretches of time? Why seek a common origin of hereditary mole- 
cules? Why do science at all? 

Science and Questions of Ethics. Granted, science has its limits. We 
expect benefits to flow to society from the scientific enterprise; but 
there are questions that science cannot be expected to answer. One re- 
cent attempt to invoke scientific evidence to answer ethical and moral 
questions merits special mention. It is no less repressive than crea- 


tionism, although possibly more subtle. We have seen introduced in 
the Senate of the United States a bill "to provide that human life shall 
be deemed to exist from conception." It reads, in part, "The Congress 
finds that present day scientific evidence indicates a significant 
likelihood that actual human life exists from conception." This state- 
ment embodies and expresses a dogmatic and dangerously narrow defi- 
nition of "actual human life," for human life cannot properly be said to 
begin at any single moment fixed in time. 

Indeed, human life is a continuum, proceeding generation after gen- 
eration. The eggs contained in the ovary of the young girl ripen and are 
shed over her reproductive lifetime. These eggs, like the other cells of 
the woman's body, are living. The sperm maturing in the human male 
are no less alive. The union of living egg and living sperm results in a 
living zygote, no less alive than its parental predecessors, but differing 
from both of them. The zygote is but one fleeting morphologic and 
physiologic entity in the panorama that is human development. When 
does "personhood" begin? In my opinion, the question cannot be 
answered scientifically. Some might argue for the moment of concep- 
tion, others for the moment when the heart first begins to beat, or the 
face takes shape, or the brain begins to function, or even at birth itself. 
Yet some physiologic functions do not come into play until after birth; 
and, as Sir Peter Medawar has written, "Birth is a moveable feast in 
the calendar of development." 

Nor must one be confined to the study of normal development to 
question any definition of "human personhood as beginning at concep- 
tion." The study of birth defects similarly reveals that some defects 
are indeed embodied in the genes and present from conception, while 
others are extragenetic, arising later in the course of development. 

One cannot "fix" scientifically the time when "human personhood" 
begins, for development is, to use an accepted scientific term, epige- 
netic, wherein the morphologic and physiologic characters of the body 
are delineated gradually. In fact, the question of human personhood 
cannot be considered in medical and scientific terms. It embodies 
moral and ethical questions that have occupied the mind of Man for at 
least as long as Man has been recording his thoughts. In essence, 
philosophical thought is part of that continuum that is human life 

Surely the social compact between science and society, to use 
Wright's term, requires not only that scientific evidence be marshaled 
and weighed judiciously, but also that it should be used only when the 
question is one that the methods of science can reasonably be expected 
to answer. Thus, I would hope that such difficult and important ques- 
tions as the performance of abortions, whether or not at public ex- 
pense, would not be regulated arbitrarily by invoking "present day 


scientific evidence" that "actual human life exists from conception," 
when the circle of life is, in fact, continuous. 

These examples, two among many, speak to the breakdown of the 
link between the political and the scientific aspects of social greatness, 
a link, as Haskins has cogently pointed out, so fundamental that even 
its nature is not always explicitly recognized. I would be less concerned 
if I thought that this breakdown were an aberration springing from 
and affecting only our own generation. Unfortunately, in Fred Hech- 
inger's words, "The science age is turning darker. . . . This may be the 
age of science everywhere except in American public schools." The 
problem is likely to be exacerbated, for contemporary American soci- 
ety lacks a vision of the value and role of education. 

The Value and Role of Science Education 

... it is depressing to try to write a book while worrying 
whether anyone will take time to read it, but also because my 
research yielded conclusions so much more disheartening and 
sad than those I had anticipated. 

Arthur Levine 

When Dreams and Heroes Died 

A Portrait of Today's College Student 

As John A. Moore has written, "... the current success of the crea- 
tionists is the most severe indictment that can be leveled at the 
teaching of science in the schools and universities of America." It 
reveals the shocking scientific illiteracy of our citizenry. 

The United States has not made an explicit commitment to provide a 
grounding in science and mathematics for most of its citizens. There 
are problems fundamental to all levels of science, mathematics, and 
engineering education— deeply disturbing trends that, if allowed to 
continue, threaten the long-term vitality of the scientific and technical 
endeavor, and ultimately, the nation's well-being. 

The current state of affairs has not come about suddenly, nor is it 
new. It reflects a long-standing indifference to the systemic aspects, 
for science and mathematics education are hierarchical in character 
and are especially vulnerable to discontinuity in the learning chain. 
The educational foundations gained in elementary and secondary 
school are vital. Among problems at these levels are a relatively fixed 
teaching corps— troubled by deficiencies in training and professional 
stagnation, struggling with outdated curricular materials and instruc- 
tional equipment, often working with increasingly disinterested and 
distracted students. As we have already observed, the state of scien- 
tific instrumentation, for teaching as well as research, is deplorable; 


there is no other word for it. Yet, despite the gravity of our situation, 
we find no evidence of coherent actions and sustained leadership from 
the federal government. 

As I have remarked on an earlier occasion, the federal role in science 
and engineering education has never been defined satisfactorily nor 
delineated with precision. Neither the Department of Education nor 
the National Science Foundation has afforded a clear vision of priority 
and purpose; neither has provided the continuity of funding needed for 
stable programs with capacity for change. I find our failure at the 
federal level especially disheartening, for a unique amalgam must exist 
between the federal government and local communities. The national 
government must strengthen this link by developing long-range plan- 
ning, policy research, and policy formulation capabilities for science, 
mathematics, and engineering education, vital functions which do not 
now exist. 

Accordingly I must renew my plea that the Executive Office of the 
President establish, in accordance with existing statutes, a Council on 
Education in the Sciences and Engineering. Members of the Council, 
serving in a full-time capacity, should have substantial responsibility 
for developing analyses and advice that will assist the President in the 
shaping of long-term national and federal policies and priorities for 
science and engineering education. 

Moreover, in the face of the creationism movement, I urge that cur- 
riculum renewal again be made an ongoing sustained effort. For the 
very essence of science and technology is change. Ways must be found 
to update and develop science and mathematics curricular materials at 
both primary and secondary levels. Distinguished scientists and 
teachers must work together. Opportunities to inform and acquaint 
teachers at all levels with these materials should be an integral part of 
all curriculum development projects, and mechanisms must be devel- 
oped to ensure periodic review and renewal of materials. 

The continuing education of teachers and the innovation and renewal 
of instrumentation would be facilitated by the development of more 
local and regional, science and mathematics teaching and resources 
centers, to work at both the elementary and secondary levels. 

These recommendations incorporate some of the principal lessons 
learned from earlier periods in curriculum development. Although it is 
commonly said that federal efforts in curriculum development came 
after Sputnik I, the seeds of change had been planted at least half a 
decade earlier. 

The late 1940 's were a time of ferment in the sciences. Extraordinary 
advances were astir in almost every field. Moreover, nearly all fields 


were populated by mature, strongly motivated investigators who had 
returned from their responsibilities during World War II, often fa- 
vored with war equipment ready to be adapted to peacetime uses, with 
ideas arising out of advances in wartime technology, and with con- 
fidence that the federal government would exert an active role in the 
advancement of science. Physics was in the spotlight, of course, but 
biology, owing to the emergence of the one gene-one enzyme hypothe- 
sis, was fast coming into its own. The revolution in molecular biology 
loomed just ahead. Thus, in 1950, research had reached high status 
amid a climate of great change. 

These matters were not lost on college and university professors, 
who were themselves riding the crest of a wave of public acclaim. 
Although higher education received the greatest attention, none 
doubted the overall role of education, including elementary and second- 
ary education, in achieving national goals. But even in the best of 
times, the fruits of research enter the educational pipeline slowly, and 
the extraordinary changes in research wrought by the war had not yet 
trickled into the educational system. 

The pioneering efforts in curriculum reform in elementary and sec- 
ondary schools were in mathematics; a curriculum was developed by 
the University of Illinois Committee on School Mathematics in 1951, 
bolstered by a grant from the Carnegie Corporation in 1952. In phys- 
ics, the Physical Science Study Committee was funded by the National 
Science Foundation beginning in 1956. After Congress passed the Na- 
tional Defense Education Act of 1958, efforts followed in chemistry, 
biology, mathematics, and earth sciences. These later programs dif- 
fered from the earlier ventures in that there was greater tendency for 
college and university scholars to work closely with their counterparts 
in the high schools. Even so, at least throughout the early years, there 
was a current of resentment arising from the feeling that the "dogma" 
was being handed down to the high schools. Gradually, this source of 
criticism lessened as more and more high school teachers participated 
in testing and revising the products of the studies. 

Changes in curriculum were most evident in the new textbooks, for 
the older books were seen by nearly all concerned (at every level) to be 
unsatisfactory. There was too little of the scientific method in the 
older, customary ways of teaching science; there was greater emphasis 
on the memorization of facts than on understanding the relationship of 
scientific theory to observation, experiment, and discovery. 

In retrospect, we see these programs to have been successful. Some, 
notably the Biological Sciences' Curriculum Study, had a profound im- 
pact on the teaching of biology in the high schools. There appeared to 
be a national consensus that federal involvement in education— at least 
in science and mathematics— was desirable and, indeed, necessary. 


Gradually, however, as agreement on broad national purposes dissi- 
pated, special interests became more prominent and curriculum devel- 
opment lost its priority. The National Science Foundation lost interest, 
especially after its curriculum project MACOS {Man— A Course of 
Study) generated public outcry. But the federal role in education did 
not decrease— rather its shape was changed. Concern was redirected 
from the quality of education to the promotion of wider access for 
groups of persons that had not participated equally in opportunities 
for education, certainly itself a worthy goal. 

My concern is with the teaching of mathematics and science to all 
elementary and secondary students, not only to prospective scientists 
and the "severely gifted"— and not only in the schools. Scientific and 
technical literacy must be enhanced throughout the population, but 
education for innovation requires more than schools can offer. In 
Learning Environments and Innovation, we read that "learning takes 
place at diverse times and in odd places— watching television, reading 
a book, visiting a museum .... Moreover, as the pace of technological 
change accelerates, more training occurs on the job or in a corporate 
environment where the skills are known and the new knowledge gener- 
ated." Industry has a larger and larger stake in science and technology 
education. To that end, I urge efforts that foster interaction between 
industry, the schools, and the wider community. 

As I wrote in the preface to the report of the Committee on National 
Educational Policies in Science and Engineering,* the educational 
problems that are resulting in an erosion of the dominant position of 
American technology and industry are not partisan problems. The 
seeds of decline are not rooted in any one administration nor are the 
solutions likely to emerge full-blown during the tenure of any one 
leader. We must mobilize our educational resources in a focused and 
sustained national effort, looking toward a society that will have, in 
Moore's words, "an effective understanding of the nature of science 
and its strengths and limitations . . . and a similar understanding of the 
nature of religious and humanistic statements and their strengths and 


Last year, this section of my report was entitled "Highlights of the 
Year." As always, it was intended not to render a complete account of 
the year's research, but to convey a sense of the directions the several 

* Although I have drawn upon the Report, and have had the benefit of discussions 
with members of the Committee and its staff headed by Wendell Williams and Sharon 
Bush, the opinions I have expressed are my own. 


departments are taking, and of their spirit and vitality, by singling out 
a few discoveries deserving special mention. Unfortunately the custo- 
mary disclaimer of infallibility was omitted, and recent events have 
shown how difficult it is for an embryologist to select highlights in 
fields far removed from his own. 

In Year Book 79, pp. 620-621, under the heading "The Distribution 
of Galaxies in Space," Stephen Shectman reported the discovery by 
Drs. Robert Kirshner of the University of Michigan, Augustus Oemler, 
Jr., of Yale, and Paul Schechter of Kitt Peak National Observatory, 
and himself, of a large volume of the universe "conspicuously empty of 
galaxies." I found the observation intriguing, but Shectman's straight- 
forward account did not give me a sense of its importance. Some 
months later, I read in the nation's leading newspapers the startling 
headline, "Vast 'Hole' in Space Appears to Defy Theories." Indeed, the 
discovery by Shectman and his colleagues was said to demand revision 
of current theories on how the universe evolved. 

Does the group's discovery truly challenge the generally accepted 
view of the large-scale homogeneity of matter soon after the Big Bang? 
Shectman believes that it does not, and he offers other possible ex- 
planations. But first let us review the facts. 

Most astronomers have believed that the galaxies were rather uni- 
formly distributed during the early millenia of the universe. With time, 
gravitational interactions caused many galaxies to clump together into 
clusters. The unevenness, or "dumpiness," of the universe seen today 
is an important feature confronting scientists who study the history 
and nature of the universe; all new theories must conform to the known 
dumpiness of matter. 

The nearly empty region reported by the group lies across 30-40 
degrees on the sky, and spans a diameter 100 times the distance to our 
nearest major galaxy, Andromeda. Its discovery came about in the 
following way. 

The group began their investigation of the large-scale distributions 
of galaxies several years ago. Their method was to study particular 
regions of sky by observing in detail fields of about two square 
degrees. By obtaining redshift (i.e., distance) information spec- 
troscopically, they in effect constructed three-dimensional maps of 
each field. When they examined three fields in the direction of the 
north galactic pole (avoiding the dust obscuration of the galactic 
plane), they found remarkable and similar patterns of galaxy distribu- 
tion. Of more than a hundred galaxies observed within the three fields, 
only one galaxy lay within a wide middle-distance range (correspond- 
ing to redshift values between 12,500 and 18,500 km/sec). Statistically, 
the galaxy population at these distances should have been 25. The 
nearly empty region apparently has a volume three times that of the 


largest previously suspected region of void and about 100 times that of 
any gap previously studied in this detail. 

According to Shectman, one possible explanation is that the void 
may have begun as a region only slightly depleted in matter. Such a 
region would tend to grow and become more empty because its lower 
mass density (in comparison to neighboring regions) would retard ex- 
pansion of the volume only weakly. Also, the surrounding regions may 
have depleted the gap of its mass by gravitational actions. This 
possibility is supported by the higher-than-average galaxy population 
just outside the empty region. Another possibility is that mass is pres- 
ent but in nonluminous forms— perhaps in the form of gas or very 
small galaxies. Shectman and his colleagues are continuing to in- 
vestigate the question by studying additional fields within and around 
the apparent empty region. They have also begun a similar study of 
three similar fields in the south galactic pole region with the 2.5-meter 
telescope at Las Campanas. The results should aid in interpreting the 
growing data set from the northern region. 

Structure, Dynamics, Composition, and Evolution of Galaxies 

At the risk of being accused of favoritism— having already drawn 
upon the work of Vaughan and Shectman and their collaborators in il- 
lustrating the excellence of the year's work, I will continue with 
astronomy, where the Institution's human resources are especially 

As have other directors before him, George Preston of the Mount 
Wilson and Las Campanas Observatories extolls the virtues of our 
observatory at Las Campanas, Chile. He points out that among the 
many blessings it has bestowed on astronomy, Las Campanas provides 
access to the Magellanic Clouds and to our galactic center, "which 
passes in full glory through the zenith at Las Campanas." Such advan- 
tages have reinforced an already strong commitment by the Obser- 
vatories' staff to studies of the structure, dynamics, composition, and 
evolution of galaxies, including our own. And, Preston goes on, "Just 
as biologists would be entranced by the opportunity to study the forms 
and evolution of living organisms on another planet, so are astrono- 
mers captivated by the opportunity to make analogous studies of stars 
within regions of space very different from the solar neighborhood." 

Studies of the chemical composition of various classes of stars in the 
Magellanic Clouds may provide clues to the chemical evolution of 
galaxies other than our own. Eric Persson, collaborating with Judith 
Cohen of the California Institute of Technology and Jay Frogel and 
Jonathan Elias of the Cerro Tololo Inter-American Observatory, has 


made infrared observations of more than 100 luminous red stars in the 
Clouds. Nearly all carbon stars in their sample have luminosities lower 
than those predicted by theory. The group is now considering possible 
explanations for this discrepancy. Cohen is also determining heavy- 
element abundances in the giant stars of clusters in the Large Magel- 
lanic Cloud by analysis of spectra obtained with the du Pont telescope. 

In parallel, ages and compositions have been estimated from inte- 
grated colors for a number of globular clusters in the Large and Small 
Clouds by Leonard Searle and Horace Smith. They report that the 
Large Magellanic Cloud contains many globular clusters with ages be- 
tween 2 and 10 billion years, much younger than the globular clusters 
of our own Galaxy. The fact that these clusters resemble the intermedi- 
ate-age, metal-poor open clusters of the galactic anticenter leads Searle 
and Smith to conclude that the chemical history of the Large Magel- 
lanic Cloud probably resembles that of the outer fringes of the galactic 
disk rather than that of the solar neighborhood. Graduate students 
Matthew Malkan and R. Michael Rich have also been working at Las 
Campanas, making a related infrared metallicity study of globular 
clusters near the galactic center. 

Turning to more distant galaxies, Vera Rubin, Kent Ford, Norbert 
Thonnard, and their colleagues, whose "home base" is the Department 
of Terrestrial Magnetism (DTM), have continued their systematic 
study of the variation of rotation velocity with distance from center of 
spiral galaxies. As DTM director George Wetherill states, the velocity 
at a given radial distance for a spiral galaxy can be related in a simple 
way to the total mass lying within that distance, as a simple conse- 
quence of the laws of gravity and motion. Therefore by measuring the 
radial velocity, one can calculate the radial variation of material den- 

From the way that the luminosity of a spiral galaxy falls off with 
distance, one might expect that the galaxy's mass is concentrated 
toward the center. If so, stars in the outer regions of a galaxy would 
rotate about the galaxy center more slowly than stars near the galaxy 
center. In our solar system, for example, the innermost planet, Mer- 
cury, is moving at ten times the speed of the outermost planet, Pluto, 
because the mass of the system is concentrated almost entirely in the 
Sun. But from the work of Horace Babcock in 1939 and others, it has 
been realized that the masses of spirals are not concentrated at the 

The DTM group has explored in detail this phenomenon and has now 
completed velocity measurements on 60 spiral galaxies, including rep- 
resentatives of the three types, Sa, Sb, and Sc, classified on the basis of 
their morphology alone. 

Wetherill summarizes important generalizations already emerging 


from this ongoing study. The first, and to him the most important, is 
confirmation that the rotation velocity of spirals does not fall off with 
distance from the center. At great distances, it is either still rising or 
constant. A considerable fraction, possibly even most, of the mass, lies 
beyond the edge of the visible image. 

Another systematic property of spiral galaxies was established this 
year. At first glance, curves resulting from plotting increase in velocity 
with distance from the center are dissimilar for galaxies of different 
luminosities. David Burstein (now at the National Radio Astronomy 
Observatory), Rubin, Thonnard, and Ford show that, at least for Sc 
galaxies, these differences largely disappear when the radial distance is 
scaled properly. For each galaxy, the unit of radial distance is trans- 
formed by dividing the actual radius by the radius within which a stan- 
dard mass (10 10 solar masses) is found. When this is done, the in- 
dividual rotation curves fall on top of one another to form a single "uni- 
versal" rotation curve, on which most of the data from individual 
galaxies plot. The underlying cause of this regularity is still not well 
understood, but its discovery should aid in solving the fundamental 
problems of the structure of galaxies as they are now observed, as well 
as the manner in which this structure originates and evolves. 

In many galaxies the rotation velocity is still rising at the outermost 
edge of the optically observed galaxy. Less direct methods must be 
used to learn how much "invisible" mass lies at even greater distances. 
Rubin and Ford discuss an approach wherein a distant background 
quasar serves as a source of radiation enabling detection of the outer- 
most, invisible part of a galaxy. Previous radio astronomical work by 
other investigators had identified a narrow absorption line in the spec- 
trum of quasar 3C232, signifying a cloud of neutral hydrogen gas 
somewhat outside the visible edge of the galaxy NGC 3067. If it is 
assumed that the cloud lies in the plane of the galactic disk, then its cir- 
cular velocity and distance from the galactic center can be calculated. 
It turns out that this velocity agrees with the velocity calculated by ex- 
trapolating the small, positive-velocity gradient at the visible edge of 
NGC 3067 out to the distance of the cloud, a distance about four times 
the visible galactic radius. Although it can still be questioned whether 
or not the cloud is actually in the galactic plane, the result suggests 
that the rotational velocity continues to rise far beyond the edge of the 
optical image. If so, this implies that almost all the mass ( > 95%) of the 
galaxy is "invisible," i.e., nonluminous. If this result is typical of other 
galaxies, this would affect estimates of the mean density of the 
universe, and would provide important evidence as to whether or not 
the observed expansion of the universe will continue indefinitely. 

Wetherill also reviews a study of elliptical galaxies by Brad Whit- 
more and his colleague Robert Kirshner at the University of Michigan 


(who has also collaborated with Shectman). This study focuses on the 
similarities between elliptical galaxies and the central bulges of spiral 
galaxies, and asks whether it is appropriate to consider a spiral galaxy 
as an elliptical galaxy plus a disk containing spiral arms . To approach 
this question quantitatively they had to obtain accurate luminosity 
profiles of the central bulges of spiral galaxies, distinct from the disk 
components. Earlier photometric work of this kind has given inconsis- 
tent results. In this new work, more accurate photoelectric techniques 
replace the photographic methods previously used to determine bright- 
ness. So far, the main effort has been to establish the accuracy of the 
present and earlier photometry. Ultimately, accurate surface bright- 
ness profiles of spiral galaxies will be combined with mass profiles of 
the same galaxies in order to determine local values of the mass-to- 
luminosity ratio. This information will be fundamental in understand- 
ing the interplay of galaxy dynamics with stellar birth, evolution, and 

Similar questions have occupied several members of the staff at the 
Mount Wilson and Las Campanas Observatories. Alan Dressier and 
Allan Sandage have also been interested in the relations of elliptical 
and spiral galaxies. They have addressed the question whether dif- 
ferences in galaxy type are due to evolutionary change or rather to dif- 
ferent initial conditions at the time of formation. To this end, they ob- 
tained rotation curves of 25 SO galaxies (galaxies with stars in a flat 
disk and spheroidal bulge but with no gas, dust, or recent star forma- 
tion). Although SO's possess some of the morphological characteristics 
of spirals, they appear to be a transition between disk-dominated 
spirals and ellipticals. The rotation measurements of their study are 
confined to the inner few kiloparsecs of the bulge. From them, Dressier 
and Sandage conclude that the bulges of SO galaxies are rotating much 
more rapidly than elliptical galaxies of comparable shape and size. This 
fundamental dissimilarity implies that the initial conditions for the for- 
mation of elliptical galaxies and the bulges of disk galaxies must have 
been different. 

As part of the Palomar-Groningen Survey of Bright Galaxies in pro- 
gress for the past three years, Leonard Searle and Pieter van der Kruit 
have completed a study of two bulge-free, edge-on spirals, NGC 4244 
and NGC 5907, comparing their three-color surface photometry of 
these systems with a model characterized by an exponential radial 
decline in surface brightness and a light distribution perpendicular to 
the symmetry plane appropriate for a self-gravitating isothermal 
sheet. They found that an excellent fit of model predictions and obser- 
vations could be obtained by requiring that the perpendicular scale- 
height be independent of galactocentric distance and that there be a 
boundary galactocentric distance beyond which stars do not form. 


In a related study of NGC 891, an edge-on spiral regarded as similar 
to our own Galaxy, Searle and van der Kruit found that a one- 
component disk model could not account for the observed light distri- 
bution. A second, approximately isothermal component of higher 
kinetic temperature is present. The second component is relatively 
bluer, which the investigators interpret as an abundance effect. In ap- 
plication to our own Galaxy, they predict that at 2-3 kpc above the 
plane the abundance distribution should be bimodal. A photometric 
program in progress by Sandage should provide an early test of this 

The Earth and Planets: Meteorites and Meteoroids 

This year Douglas ReVelle, now at Northern Arizona University, and 
George Wetherill completed an extended study employing data from 
several recovered meteorites. In introducing the report, Wetherill 
points out that almost all meteorites, which are the oldest rocks we can 
ever expect to find, were assembled from small fragments or were 
crystallized from magmas during the formation of the solar system. 
Over the years, laboratory studies of meteorites have provided detailed 
and accurate data on primordial chemical and isotopic compositions, 
mineralogy, radiation exposure, and many other measurable quan- 
tities. There is a difficulty, however, in distinguishing the essential 
from the incidental in using these data, namely the absence of geolog- 
ical context. 

Wetherill finds reason to hope that this problem will not be perma- 
nent. In comparisons with lunar samples, powerful analogies are being 
drawn between meteoritic and lunar breccias. These analogies provide 
good reasons for believing that certain types of meteorites are derived 
from airless bodies similar to the Moon but smaller and more distant— 
the asteroids. Advances in astronomical techniques, particularly in the 
fields of infrared spectroscopy and orbital dynamics, have even en- 
couraged some investigators to identify particular kinds of meteorites 
with particular asteroids. 

Progress in understanding meteoritic origins has been made during 
the past 20 years as a consequence of recovering on the ground three 
meteorites with astrometrically determined orbits. ReVelle and Weth- 
erill have used data from these recovered meteorites— in WetheriU's 
words— as "ground truth" for identifying similar, but unrecovered, 
meteoroids with astrometrically determined orbits photographed by 
workers at the Smithsonian Astrophysical Observatory. This work has 
resulted in a list of 27 such objects that are indistinguishable from the 
recovered meteorites in their physical properties while passing 
through the atmosphere, expanding the catalog of known meteorite or- 


bits tenfold. It should provide a much better basis for comparing the 
predictions of dynamic theory with observation. There exist other 
unreduced data of a similar kind, and Wetherill believes that it should 
be possible to obtain about 100 well-certified meteorite orbits from ex- 
isting photographs. 

When a sufficiently large set of data becomes available, it may 
become possible to make finer distinctions among solar system sources 
of meteorites of various types (e.g., stony vs. iron meteorites). When 
combined with geochemical data, information from the meteorite 
studies could provide insights into the physical and chemical condi- 
tions in the early solar system as a function of distance from the Sun. 
For example, Brajesh Kothari, in collaboration with Peter Maggiore 
and Edward Scott at the University of New Mexico, report plutonium 
fission track measurements and metallographic cooling rate data for 
an unusual type of stony meteorite. Chemical data indicate a relation- 
ship between these, meteorites and silicate inclusions in a major type of 
iron meteorite. The goal of reconstructing the penological histories of 
the asteroidal sources of meteorites such as these is still distant. Its 
achievement will almost certainly require the use of data from all 
available sources: fireball observations, theoretical dynamics, and 
detailed geochemistry and petrography. 

A principal reason for associating meteorites with asteroids is the 
fact that even though the meteorites are 4.5 billion years old, careful 
radiometric dating of meteorite samples has provided evidence of later 
isotopic disturbances, in particular disturbances that appear to be 
associated with collisions. It can easily be calculated that high-velocity 
collisions are common between asteroids, but are very improbable 
among comets. On the other hand, disturbances caused by aqueous al- 
teration could possibly occur in comets. 

Fouad Tera has described a critical examination of published iso- 
topic U, Th, and Pb studies of meteorites, with particular attention to 
evidence for recent chemical disturbance of these isotopic systems. 
Such evidence is ubiquitous, he finds, and he concludes that if the 
published data are taken at face value, almost all meteorites have ex- 
perienced major chemical disturbances in their U, Th, and Pb concen- 
trations. These disturbances are sufficiently recent, however, to permit 
combined use of the 238Tj_206p D ^d 235jj.207p D a g e determination 
methods to infer the original 4.5-billion-year age of the meteorite. 
Because these disturbances are recent, it is possible that they are pro- 
duced in the laboratory, i.e., by major analytical errors. Most of the 
measurements were made by highly experienced and careful scientists, 
so this explanation may not be the correct one. If it can be shown that 
these disturbances are not laboratory artifacts, further studies of the 


origins of the disturbances could provide clues for identifying meteor- 
ites with their parent sources. 

There are, however, some phenomena observed in meteorites for 
which parent-source identification may not even be very relevant. A 
number of carbonaceous meteorites contain inclusions consisting pre- 
dominantly of Ca, Al, Ti, and Mg oxides and silicates. These inclusions 
are far out of chemical equilibrium with the rest of the meteorite. Some 
evidence for alteration exists. But to a large extent these inclusions 
can be thought of primarily as samples of the earliest solar system, 
material caught up by a larger body 4.5 billion years ago and brought 
by a similarly incidental chain of events to Earth. The unidentified 
asteroidal or cometary parent body of the meteorite can to some extent 
be thought of as merely a transportation vehicle whereby these primi- 
tive fragments were delivered to the Earth and subsequently to the 

Because of the importance of this material as a relic of early solar 
system history, these samples are being studied by many workers. 
Sundar Raj an and Arvind Tamhane have studied fission tracks in hibo- 
nite from the Murchison carbonaceous chondritic meteorite. (Hibonite 
is a very refractory calcium-aluminum oxide that has been interpreted 
by many workers as the very-highest-temperature condensate from the 
pre-planetary solar nebula.) Raj an and Tamhane conclude that these 
tracks— the first to be revealed in hibonite— were almost all caused 
by fission of the now-extinct radioisotope (82-million-year half-life) 
244 p u when combined with their measurements of the U concentration 
in these same samples, their data permit calculation of the Pu/U ratio 
present 4.6 billion years ago. These ratios vary considerably, in agree- 
ment with Mg isotope, geochemical, and petrographic evidence that 
these high-temperature inclusions in Murchison constitute more than 
one population. When such results are combined with other forms of 
evidence, it may be possible to characterize these populations more 

Wetherill notes that high-temperature inclusions in carbonaceous 
meteorites (especially the Allende meteorite) are probably most famous 
as the carriers of isotope anomalies, anomalies that may be primarily 
relics of pre-solar system nucleosynthetic heterogeneities— a link be- 
tween laboratory geochemistry on the one hand and theoretical and 
observational astrophysical studies of star formation and stellar evolu- 
tion on the other. 

A perplexing aspect is that most inclusions exhibit only oxygen 
isotope anomalies. There is as yet no petrographic or geochemical 
distinction between these common inclusions and those with a variety 
of anomalies. Typhoon Lee reports finding two new isotopically anoma- 


lous inclusions that exhibit Mg isotope fractionations intermediate 
between the most common nonfractionated inclusions and the rare, so- 
called "FUN" inclusions, which contain a large number of isotopic 
anomalies. The presence or absence of these various anomalies and 
their correlations or lack thereof continue to be major puzzles in 
planetary science. 

Geochemistry and Seismology at DTM 

In his essay this year, George Wetherill pays particular attention to 
the formation of volcanic rocks, which are, he says, major products of 
the "huge chemical factory" that is the Earth— "continually engaged 
in processing and reprocessing itself." These rocks, derived from the 
Earth's interior, present targets of opportunity par excellence for 
geochemists, for they provide windows to the past, to the geological 
processes and events of the earlier history of the Earth. Interpretation 
of this record is far from simple, because the melted rock, or magma, 
receives chemical contributions from a variety of sources during its 
migration from the interior. We must learn more about the relative im- 
portance of these chemical sources: various regions in the Earth's man- 
tle, recycled volcanic and other crustal igneous rocks, and sediments 
derived from oceanic and continental regions and subducted back into 
the mantle along with the oceanic lithospheric slabs upon which they 
were deposited. 

One group of investigators has made remarkable progress in devel- 
oping a technique for using the naturally occurring radioisotope 10 Be 
to quantify the contribution of subducted sediments to lavas erupting 
at the surface. The technique employs a Van de Graaff accelerator as a 
sensitive and selective mass spectrometer. Meanwhile, Robert Stern 
addresses the origin of island-arc volcanic rocks by geochemical and 
isotopic comparisons of volcanic rocks from two regions of the Pacific- 
one lying above a subduction of an oceanic plate, the other within an 
oceanic plate and thus not associated with a downgoing slab. Stern 
finds that the regions are similar in geochemical and isotopic proper- 
ties, and he proposes that island arcs and oceanic islands may be de- 
rived from a common mantle source, one less depleted in the incompati- 
ble elements than is the source of ocean ridge basalts. 

About a decade ago, soon after the general acceptance of the plate 
tectonic synthesis, there was a tendency to regard the continents as 
superficial "scum" riding along on the backs of several of the great 
lithospheric plates. The plates were believed to be the true building 
blocks of the Earth's outer regions. We are gradually realizing that the 
differences between oceans and continents are not confined to the 


shallow crustal regions above the continental Mohorovicic discontinu- 
ity, at a depth of about 35 km. For example, seismic data provide 
evidence that at least some portions of the continental lithosphere ex- 
tend to depths of from 200 to 400 km, much thicker than the from 50 to 
100 km deep oceanic lithosphere. Wetherill points out that our under- 
standing of the deep structure of continents is in a very rudimentary 
state, and he writes that the opportunity to participate in many impor- 
tant discoveries awaits those brave enough to try to untangle the com- 
plexities of continental crustal and mantle geology. 

The Andean region of western South America has received con- 
siderable attention by scientists, including those at DTM. It is a very 
important area, because it represents the only presently active junc- 
tion between an oceanic plate and a deep continent. The Andean vol- 
canoes, rising above a plateau itself four kilometers above sea level, are 
an obvious manifestation of this interplate interaction. The extent to 
which there is chemical interaction— i.e., the exchange of material be- 
tween the overriding continental plate and the subducting oceanic 
plate— is much less clear. If such exchange occurs, plate junctions of 
this kind must have been an important link in the fundamental geo- 
chemical evolutionary process transferring chemical elements from the 
oceanic mantle to the continental platforms. 

Some source beneath the Andes must be found that contributes 
strontium with a high and variable 87 Sr/ 86 Sr ratio. In 1978, Mordeckai 
Magaritz and co-workers at DTM placed an important constraint on 
the source of this radiogenic strontium by obtaining 18 0/ 16 data that 
clearly indicated the involvement of water with an atmospheric his- 
tory. This result argues that a mantle source cannot entirely explain 
the origin of the magma and shows that a crustal contaminant of some 
kind is responsible for the introduction of the nonmantle ls O/ 16 and, 
one might expect, for the excess radiogenic strontium as well. 

Simple geochemical models for the source of this crustal component 
have subsequently been proposed by DTM scientists and other in- 
vestigators. Although all these models have some attractive features, 
none of them explain satisfactorily the quite extensive geochemical 
and isotopic data that became available during the pursuit of the prob- 

David James now describes the application of a more complex but 
nevertheless plausible model that other workers have found to be satis- 
factory in other regions. This model involves a chemical reaction be- 
tween a primary magma and previously solid rocks in the upper crust, 
a process that could occur in a shallow magma chamber. The solid wall 
rocks of the magma chamber melt and are assimilated by the magma, 
which at the same time is partially solidifying. The chemical and 
isotopic composition of the magma is thereby altered both by the addi- 


tion of wall rock and by the subtraction of precipitating crystals. The 
volcanic rock actually erupting from the volcano is the residual liquid 
remaining after this chemical processing has occurred. James, follow- 
ing recent work by George Tilton and B. A. Barreiro, identifies the wall 
rock with the Charcani gneiss, the basement rock underlying the vol- 
canics in the Arequipa region. Quantitative calculations show that this 
model brings unity to much of the diversity of the geochemical data, 
and permits characterization of the primary magmas prior to the late- 
stage assimilation of upper crustal material. 

These primary magma compositions no longer bear the clear 18 0/ 16 
signature of crustal contamination, and thus more nearly represent the 
partial melting product of their source. The compositions of the 
primary magmas and their source are distinctly different from those 
associated with normal island-arc volcanism, which do not involve a 
continental plate. Although much remains to be learned before the 
history of this complex plate interaction is even moderately well 
understood, this plausible way of correcting for the upper crustal con- 
tribution should open investigation of more profound questions. 

Crustal Deformation. Most of our knowledge about the structure of 
the Earth's deep interior has come from the analysis of seismological 
data, obtained by measuring the elastic waves radiating from an earth- 
quake or explosion. These measurements are made at great distances 
from the source of the elastic waves, and the conclusions concerning 
the Earth's interior are fortunately insensitive to details of the phys- 
ical processes in the source region of the earthquake. On the other 
hand, the near-source events represent the real-time manifestation of 
fundamental mechanisms of geodynamics and structural geology. Near- 
source events are also very relevant to the practical problem of earth- 
quake prediction. They can be thought of as "plate tectonics in 
action"— at the opposite extreme from the broad, long-term view pro- 
vided by reconstructions of the ancient continents of Gondwanaland 
and Pangaea. This year, our seismologists report the results of near- 
source investigations of crustal deformation in seismically active 

Recent work in seismic source theory is formulated in terms of sev- 
eral source parameters: seismic moment, source dimension, stress 
drop, and radiated energy. Of these, the seismic moment is recognized 
as important because it is based only on the initial and final states of 
the faulted region and thereby is different from parameters that in- 
volve assumptions about the nature of the rupture process. 

J. Arthur Snoke, from the Virginia Polytechnic Institute, in col- 
laboration with Alan Linde and Selwyn Sacks, describes comparison of 
two methods for measuring the seismic moment of earthquakes in the 


Matsushiro region of Japan. The first is the conventional method, and 
employs the Carnegie broad-band seismograph system at Matsushiro. 
The second is a new method and makes use of static offsets recorded on 
Sacks-Evertson borehole strainmeters at the same place. Similar re- 
sults are obtained by both methods. On the average, the agreement is 
within a factor of about two, which is considered acceptable. Further- 
more, these seismic moments correlate with earthquake magnitude in 
the same manner found in earlier work. Based on these first measure- 
ments in which strain steps have been used to calculate seismic 
moments, Snoke and colleagues conclude that the volume strainmeter 
is a useful instrument for measuring this important source parameter. 

Sacks, Linde, Snoke, and Shigeji Suyehiro report the results of a 
detailed analysis of the 1978 Izu slow earthquake data described last 
year. Slow earthquakes, or "slowquakes," are similar to normal earth- 
quakes in that they are episodic, in contrast to continuous plastic 
deformation. They differ from normal earthquakes in that their time 
scale is of the order of an hour, rather than only seconds. As a result of 
the longer time scale, the accelerations associated with slowquakes are 
much less and physical damage is minor. On the other hand, these in- 
vestigators suggest that the more subtle slowquakes may constitute 
an important, but heretofore unrecognized, mechanism for distributing 
the crustal stresses that determine the locations of damaging earth- 

The present analysis considers the strain changes recorded by three 
borehole strainmeters following the 1978 normal Izu earthquake (mag- 
nitude 7.0). These strain changes are interpreted as a slowquake se- 
quence, which redistributed the stress that after more than an hour 
caused a series of normal aftershock earthquakes centered at distances 
up to 30 km from the main earthquake. 

In their analysis, the region is modeled according to elastic deforma- 
tion theory, and the strain field is calculated as a function of time, con- 
strained by the aftershock data and the observed fault motions. The 
calculated displacements agree with geodetic data obtained from level- 
ing surveys and triangulation. The calculated strains and those ob- 
served by the strainmeters are also in good agreement, implying that 
the suggested relationship between the observed strain changes and 
those responsible for the fault displacements and aftershocks is valid. 
The rupture velocity along one of the faults is also determined and is 
found to be only 0.1 km/sec, much slower than that found for fault 
breaks associated with normal earthquakes. The total seismic moment 
associated with the slowquake sequence is about 40% that of the main 
normal earthquake. This work strongly supports the inference made 
earlier as to the existence of large slowquakes and their role in stress 


The Earth Sciences as a Frontier Institution 

Geology evolved as a mechanism for coping with certain in- 
tellectual problems, particularly the consciousness of a land- 
scape whose spatial and temporal scales were rapidly expand- 
ing. Prior to the latter half of the eighteenth century, western 
civilization's geographic knowledge was limited to the coast- 
lines of most of the world's continents. The only continental in- 
terior known with any precision was that of Europe. 

For America, moreover, geology had additional importance as 
a frontier institution, as economic and intellectual subsidy to 
the westward migration. The discovery of landforms, rich new 
soils, lodes of precious minerals, water resources, breathtaking 
vistas and scenes of high adventure coincided with an outburst 
of cultural nationalism and a sprawling folk migration across 
North America. 

Stephen J. Pyne 

Grove Karl Gilbert 

A Great Engine of Research 

The earth sciences are no less a frontier institution today, and no less 
crucial to our national vitality and security. Today, however, the breath- 
taking vistas are not of the western landscape, but of the behavior of 
matter at high pressure, of element partitioning, of the chemistry of 
crystals, and of the application of biogeochemistry to paleontology, 
stratigraphy, and economic geology. These are just a few of the direc- 
tions being taken at the Geophysical Laboratory. 

It was exactly 75 years ago that ground was broken for the Geophys- 
ical Laboratory. Since then, in their fundamental research on the 
chemistry and physics of the Earth, the Laboratory's scientists have 
provided a groundwork of principles later applied by society in solving 
emergent problems. Today, as we foresee critical needs for new mineral 
and energy sources, the work of the Laboratory is more vital than ever. 

Biogeochemistry. There are today countless exciting research vistas 
in the application of biochemistry to earth science problems. Paleon- 
tology has been expanded by the discovery of residual, indigenous, 
organic material in fossils. Stratigraphy has been given new tools for 
correlation, even where index fossils are absent or morphologically in- 
determinate. The petroleum industry has received benefits from bio- 
geochemistry, and economic geology has acquired new insights from 
the interactions, passive or active, of ore solutions and organic mate- 
rial. Even biochemistry itself has been enriched by the new techniques 
demanded by earth scientists for dealing with minute amounts of com- 
plex organic assemblages. In short, biogeochemistry is at that pioneer- 
ing stage where every experiment is a new challenge and yields a new 


array of interesting possibilities. For example, reports of D-enan- 
tiomers of amino acids in diseased tissue suggest that the techniques 
developed for biogeochemical investigations will eventually help 
resolve some health problems. 

As director Hatten Yoder's commentary reveals, workers at the Geo- 
physical Laboratory enjoy the freedom to focus on any substantive 
problem without regard to discipline boundaries; thus it is not surpris- 
ing to find that one current investigation involves thermophilic algae 
and photosynthetic bacteria— unusual subjects for earth scientists. 

The thermal springs of the western United States are inhabited by 
mat-forming, blue-green algae and filamentous photosynthetic bacte- 
ria—organisms possibly similar to those important in the formation of 
certain Precambrian fossils (stromatolites). Marilyn Estep has mea- 
sured the stable carbon and hydrogen isotopic compositions of several 
mats collected from hot springs in Idaho. Blue-green algae have a 
distinctly different isotopic composition from the photosynthetic 
bacterium Chloroflexus, which at higher temperatures (55 °C) forms 
pure mats. When Chloroflexus grows in association with unicellular 
blue-green algae, however, its isotopic composition shifts to one ap- 
proaching that of the blue-green algae. This shift indicates a change in 
the bacterium's metabolism from photosynthesis to photoheterotro- 
phy; in essence, the bacterium grows by using photosynthetically pro- 
duced organic matter from the blue-green algae. Even though the 
organic carbon and bonded hydrogen can be transferred from one 
species to another, the isotopic composition of these algae and bacteria 
cannot be equated with that of the Precambrian stromatolites. In addi- 
tion, no known nonbiological process produces the required shift in 
isotopic composition. Other causes for the formation and isotopic com- 
position of the stromatolites must be sought. 

One of the major accomplishments of the past few years— the appli- 
cation of geochemistry to stratigraphic problems— has provided new 
tools to supplement classical methods. Julia Corrado and P. Edgar 
Hare have performed ion-exchange, amino acid analyses on approx- 
imately 180 individuals of the pelecypod Mulinia lateralis (Say) from 
Pleistocene sediments in the Coastal Plain of South Carolina, in 
regions near Charleston and Myrtle Beach. In both regions, groupings 
of alloisoleucine/isoleucine values were indicative of relative geologic 
age. Stratigraphic units in Charleston and Myrtle Beach have been ten- 
tatively correlated on the basis of alloisoleucine/isoleucine value group- 
ings, ranges in elevation of associated Coastal Plain terraces, and 
geomorphic evidence. The number of depositional events that can be 
inferred from the value groupings and the striking similarity of the 
groupings themselves suggest a correlation of stratigraphic units be- 
tween the two regions. The ranges in elevation of the terrace features 


associated with the four value groupings in Charleston, from youngest 
to oldest, overlap almost perfectly with ranges in elevation of the ter- 
races associated with the youngest-to-oldest value groupings in Myrtle 
Beach. The sedimentary sequence, in which the next-to-youngest fos- 
siliferous unit in both Charleston and Myrtle Beach is buried in an 
almost identical fashion by the youngest unit, again suggests a direct 
correlation of stratigraphic units between these localities. With paleo- 
amino acid analyses, it will now be possible to detect regions in the 
southern Atlantic Coastal Plain that have undergone faulting and, 
more generally, to correlate beds where the fossils are not diagnostic or 
where the fossil assemblages record similar repetitive transgressive- 
regressive events. 

Fluid Speciation and Dynamics. Perhaps one of the most critical 
areas for research today is in ascertaining the components in an ele- 
ment-transporting fluid. Furthermore, it is essential to know, on both a 
local and a regional scale, the direction of flow of that fluid. Such 
knowledge underlies the entire quest for ore deposits and the under- 
standing of the processes of metamorphism. As is the custom at the 
Geophysical Laboratory, the principles are first worked out with sim- 
ple systems where the variables can be defined and controlled. The 
principles are then applied to field problems to determine whether 
other variables should be taken into account to explain the observa- 

One of several different approaches to the fluid speciation and dy- 
namics problem is that of Douglas Rumble and Thomas Hoering. 
These workers have determined that infiltration rather than intergran- 
ular diffusion through static pore fluid is the most effective mechanism 
of oxygen isotope equilibration during metamorphism. They measured 
the 5 18 of quartz separated from small samples spaced at 1-cm inter- 
vals across contacts between rocks whose 18 content differed signifi- 
cantly prior to metamorphism. In outcrops where there is no evidence 
of fluid flow during metamorphism, differences in composition of l%o 
have been found in quartz separated from adjacent samples 1 cm apart. 
Quartz from outcrops with evidence of fluid flow, such as infiltration 
metasomatic mineral zones, however, is homogeneous over distances 
of several meters. The data confirm that infiltration is a mechanism 
capable of transferring oxygen isotopes over distances greater than 1 
meter and is probably the most efficacious mechanism for oxygen iso- 
tope equilibration during metamorphism. 

Comparative Crystal Chemistry. Understanding of the physical and 
chemical behavior of a crystal depends on accurate knowledge of its 
structure. The interpretation of all properties eventually must be re- 


lated to the interatomic forces in the crystal. The crystallographers at 
the Geophysical Laboratory have provided fundamental structural in- 
formation on which to base other types of investigations seeking char- 
acterization of those interatomic forces. 

One of the most direct contributions has been in ascertaining the 
compositional limits of a crystal imposed by its general structural ar- 
rangement under a given set of conditions. For example, kornerupine, 
(Mg,Fe,Al) 9 . 10 (Si,B,Al) 5 (O,OH) 22 , is a complex silicate that, unlike many 
minerals, occurs with limited substitution of iron for magnesium. 
Larry Finger and Robert Hazen have refined the crystal structure and 
determined the site ordering of a sample with Fe/Mg = 0.5, a composi- 
tion near the most-iron-rich natural sample. Iron and magnesium are 
disordered over three of the seven octahedral and larger sites. There 
are no obvious structural limitations on the maximum iron content, at 
least under the conditions of room temperature and pressure. The 
mineral coexists with other iron-bearing phases such as biotite, cor- 
dierite, garnet, and possibly alkali feldspar, so it would appear that the 
iron content of kornerupine is governed by the partitioning coefficients 
rather than by the structure. In short, the relative free energies of all 
sites where Fe may exist in all coexisting phases determine the popula- 
tion of iron ions in each site. The calculation of the absolute free energy 
of each site, from which the relative free energies may be evaluated, is 
indeed one of the long-range goals of the crystallographer, the thermo- 
chemist, and the solid-state physicist. 

Element Partitioning. Trace element and isotopic data from ultra- 
mafic nodules in kimberlite and alkali basalt indicate that the source 
regions of these rocks may have undergone metasomatic alteration. 
Bjorn My sen has performed rare earth element (REE) partitioning ex- 
periments to determine crystal-fluid partition coefficients relevant to 
such processes. The experimental results show that even for pure C0 2 
fluids, the REE will have a preference for the fluid over any other 
phase in the upper mantle. The pressure effect is diminished, however, 
as the CO 2 content of the fluid increases. My sen showed that fluid- 
mineral partition coefficients increase rapidly with decreasing pressure 
under isobaric conditions and that the temperature effect increases 
with increasing C0 2 /H 2 of the fluid. In applying these data to rock- 
forming processes, he concluded that most likely the source rock of 
alkali basalt had undergone C0 2 -rich metasomatism prior to partial 
melting at a depth equivalent to about 30 kbar. The source of the 
metasomatizing fluid could be related indirectly to the source rock of 
the alkali basalt. Alteration of the ultramafic rock from which the 
nodules in kimberlite were derived probably occurred at pressures in 
excess of 50 kbar. The source of the fluid apparently was eclogitic; it 


could not have been a garnet lherzolite, as the fluid from such a rock 
would not show sufficient light REE enrichment. 

Experiments Bearing on the Earth's Lower Mantle and Core. Condi- 
tions equivalent to the core-mantle boundary of the Earth can now be 
studied in the laboratory. When materials of the upper mantle, as 
deduced in part from the phases in nodules from kimberlites, are com- 
pressed, they undergo further reactions and phase changes. The miner- 
als produced consist of a few simple structures, closely packed and of 
extensive solid solutions. The characterization of these structures and 
their compositions, the correlation of their properties with the proper- 
ties of the core and mantle as determined by geophysical techniques, 
and an understanding of their bearing on earth processes are the prin- 
cipal goals of the experimenter. Materials examined this year include 
metals, oxides, sulfides, and arsenides. A technique has been demon- 
strated that yields the elastic properties of crystals at very high 
pressure. In addition, a critical test of the finite-strain equation of 
state of a crystalline material has been made over an extensive experi- 
mental range. These accomplishments result in many new oppor- 
tunities for geophysicists and physicists to study materials under the 
extreme conditions of the lower mantle and core. 

Ho-Kwang Mao, Guangtian Zou, and Peter Bell began an experimen- 
tal study of metallic iron and iron sulfide (FeS) at very high pressures, 
because those materials are presumed to be among the principal miner- 
als of the Earth's core. They observed the a-e transition in iron and the 
sluggish behavior of the reaction at hydrostatic pressures up to ap- 
proximately 160 kbar. In FeS (troilite) two transitions were noted (48 
and 130 kbar) that may shift the calculated FeS eutectic point so often 
used as a constraint in geophysical models of the core. The investiga- 
tors determined the equations of state of the high-pressure phase of 
FeS to 600 kbar and calculated values for the sulfur content of the core 
to be approximately 7-9 wt %. Combination of the FeS data with 
shock-wave data on FeS 2 and Feo.gS and with their previous data on 
metallic Fe (the high-pressure e phase) revealed a trend of convergence 
of the structures to a dense, close-packed configuration of the atoms. 
At pressures above 600 kbar the density is, to a first approximation, 
sensitive to composition alone. Comparison of the specific volumes 
yields a linear, additive relationship, and higher compression appar- 
ently does not result in sharp breaks in this relationship. The range of 
values of the specific volume of the Fe-S system include the seismically 
derived values of the core! The implication is that an iron-sulfur liquid 
composition is sufficiently dense and close packed to satisfy the cur- 
rently assigned properties of the core. 



A Shift in Direction 

Winslow Briggs makes special mention of a significant shift at the 
Department of Plant Biology in the direction of research on plants 
under stress. During the past decade, a major share of the Depart- 
ment's efforts has been devoted to learning how desert species tolerate 
the environmental extremes to which they are exposed. Until recently, 
these studies have concentrated on high-temperature stress. Thus Olle 
Bjorkman, Joseph Berry, and their associates have learned much 
about the kinds of photosynthetic carbon pathways in plants adapted 
to high temperature, the components to expect in a membrane that 
enable it to remain functional at high temperatures, and finally how 
those plants that can grow at wide ranges of temperatures modify their 
biochemistry in response to changing temperature— the process known 
as acclimation. 

This year, however, the Department reports relatively little on accli- 
mation, on plants from arid regions, and on problems of temperature 
stress. Instead, attention focuses on water stress and photoinhibi- 
tion— damage to photosynthetic capacity occurring when more light is 
absorbed by a leaf than can be used by normal photosynthetic reac- 
tions. Photoinhibition can be induced by any factor restricting use of 
the reducing power generated by light-driven electron transport. Low 
temperature, limitation in C0 2 availability, and water stress can in 
combination with high light intensity all lead to photoinhibition; as will 
become clear below, all of these ultimately produce similar kinds of 
damage. Since in arid regions any or all of these factors may obtain, a 
knowledge of their specific effects is crucial to complement our already 
extensive knowledge of high-temperature effects. Such considerations 
make the shift in research direction both logical and timely. 

Photoinhibition. All told, five studies of photoinhibition are 
reported. In one, Kazuhiko Satoh and David Fork investigated the ef- 
fects of anaerobiosis on the photosynthetic chloroplasts from the green 
alga Bryopsis. Chloroplasts in excellent condition can readily be ob- 
tained from this organism, making it a particularly favorable subject 
for studies of injury effects. It is known that oxygen can serve as a 
competing electron acceptor for photosynthetically generated elec- 
trons. Satoh and Fork reasoned that if this was so, oxygen could ac- 
tually act as a protectant when conditions leading to photoinhibition 
prevailed. This prediction was verified by the demonstration that even 
under very low light fluence rates, photoinhibitory damage could be 
detected if oxygen was excluded. 

Last year, Stephen Powles, Berry, and Bjorkman showed that at low 


temperatures, for example at 6°C, high light intensity led to photoinhi- 
bition in the chilling-sensitive bean, Phaseolus vulgaris. This year, 
Fork, Gunnar Oquist, and Powles investigated the effects of such pho- 
toinhibitory treatment on photosy stems I and II. Photoinhibitory 
treatment clearly led to the accumulation of inactive photosy stem II 
reaction centers. The phenomenon was detectable both with intact 
leaves (fluorescence measured from the upper leaf surface) and with 
isolated chloroplasts. However, the leaf fluorescence measurements 
showed the effect much more dramatically, indicating that chloro- 
plasts in the upper surface of the leaf are the most damaged. The ex- 
tracted chloroplasts, of course, represent chloroplasts from throughout 
the leaf, not just the upper surface. 

Bjorkman, Powles, Fork, and Oquist asked how water stress affected 
photosynthetic performance of the shrub Nerium oleander, in full 
sunlight and under shaded conditions. They found that water stress 
damaged whole-chain electron transport far more than it damaged pho- 
tosystem I electron transport, and they verified that there was a 
strong effect of water stress on photosy stem II. Fluorescence measure- 
ments indicated that water stress leads to photoinhibitory inactivation 
of photosy stem II reaction centers precisely as it does for bean under 
chilling conditions. 

The final two studies of photoinhibition took a rather different turn. 
The mobile laboratory also took a different turn, heading not for Death 
Valley but for San Mateo County Redwood Memorial Park, across the 
Santa Cruz Mountains from the Department. There, on the forest floor, 
in a deeply shaded environment, can be found a variety of small plants, 
among them the Oxalis oregana. The leaves of this species consist of 
three leaflets, each attached to a vertical petiole by its own pulvinus. 
By changing the relative turgor in different groups of cells, the pulvini 
regulate the position of the leaflets with respect to the petiole over a 
90° range. Normally, the light intensity incident on these plants is 
about 0.5% that of full sunlight. However, occasional sunflecks pass 
over them, and may persist for an hour or more. On arrival of the sun- 
fleck, the light intensity may increase as much as 200-fold. 

Bjorkman and Powles found that when such a jump occurs, the 
leaves fold rapidly downward, exposing only a small area to the full 
sunlight. The folding response was remarkably rapid— the lag period 
was as brief as six seconds, and folding was complete within six 
minutes. Recovery after passage of the sunfleck was slower, showing 
about a 10-minute lag and requiring another 35 minutes for restoration 
of the original horizontal leaflet position. 

Light intensities of 15-30% of full sunlight are sufficient to trigger 
the folding response. Above the threshold intensities, the lag period 
becomes shorter and the angular rate of position change more rapid. It 


is the pulvini themselves that sense the changes in incident light inten- 
sity, and the effective wavelengths are in the blue. 

By making gas exchange measurements with the mobile laboratory, 
Powles and Bjorkman were able to demonstrate convincingly that the 
folding reaction forestalled photoinhibitory damage. If the Oxalis 
leaves were constrained from folding when a sunfleck passed by, sub- 
stantial photoinhibition occurred, requiring several hours for recovery. 
Curiously, the folding reaction scarcely affected photosynthesis rates. 
Photosynthesis in these shade plants is normally saturated at very low 
light intensities. Hence, even when they are folded and presenting only 
a very small area to the Sun, there is still sufficient light in the sunfleck 
to saturate their photosynthetic machinery. 

Physiological Ecology. Death Valley has not been entirely neglected. 
Jeffrey Seemann, James Tepperman, and Berry have continued work 
on the photosynthetic capacities of some of the annual plants from 
Death Valley. These plants by and large have photosynthetic capac- 
ities well above those of most temperate species: Species of Camissonia 
show the highest values recorded to date. In most cases, the very high 
photosynthetic capacity could be accounted for by unusually high 
levels of the enzymes required for fixing C0 2 into organic matter, most 
notably the enzyme that initially binds the C0 2 for subsequent 
photosynthetic steps, ribulose-l,5-bisphosphate carboxylase-oxygen- 
ase. (After years of struggling with difficult and uneuphonious acro- 
nyms for this enzyme, workers have finally suggested the nickname 
"Rubisco.") With Camissonia, however, there is another important dif- 
ference. The Rubisco from Camissonia has a specific activity signifi- 
cantly higher than that of any other species studied. Thus, not only is 
there more of it, but it is more efficient. The basis for this greater effi- 
ciency is at present unknown. 

Our physiological ecologists also report on the phenomenon of solar 
tracking. The leaves of certain species have the capacity to follow the 
Sun with great precision, hence maximizing the interception of incom- 
ing solar energy. Dov Koller has been studying solar tracking in two 
such species— Lavatera cretica and Malva parviflora— by means of a 
tracking device which permits him to sustain a given angle for incident 
light with respect to the leaf surface, and to study the tracing as a 
steady-state phenomenon. Leaves of both species can undergo reorien- 
tation at a high velocity for as much as an hour, and the rates can far 
exceed the Earth's rotation (15° per hour). Koller has recorded rates as 
high as 90° per hour, and he has found that leaves can continue to 
coast for some time in darkness after driving illumination. The stimu- 
lus itself was shown earlier by Koller and A. Schwartz to be perceived 
not by the pulvinus itself, as is the case for the leaf-folding Oxalis 


response, but rather by the major leaf veins. Some influence is then 
transmitted down the veins to the pulvinus, which is located where the 
leaf blade is attached to the petiole. Turgor changes in the pulvinus 
mediate most of the leaf movement. There is evidence, however, that 
the petiole itself can also participate; functional differentiation of 
pulvinus from petiole is hence incomplete. The leaves can reorient 
either toward or away from the stem, and both processes probably in- 
volve active transport phenomena. Leaf tracking is yet another pro- 
cess regulated by a blue light photoreceptor. 

Molecular Biology. The new focus on photoinhibition is only one ma- 
jor change in direction at the Department. Several years ago, a Trust- 
ees' Visiting Committee referred to the tools of plant molecular biology 
as highly promising but still in need of substantial sharpening. Com- 
parison of the reports from the molecular biology group five years ago 
with those in this Year Book at once indicates the dramatic extent to 
which sharpening has occurred. Already, it is evident that we can ex- 
pect a flowering of collaboration between our molecular biologists and 
our other plant biologists as the capabilities and approaches of the 
former are brought to bear on long-standing questions in plant science. 

Michael Murray and William Thompson, continuing their detailed 
analysis of the pea genome's structure, have inquired as to the nature 
of the DNA sequences in the neighborhood of those sequences that ac- 
tually code for messenger RNA. Surprisingly, the coding sequences are 
enriched in DNA regions where there is a paucity of highly repetitive 
DNA, and are depleted in regions wherein the highly repetitive ele- 
ments abound. Those repeats which co-isolate with gene sequences in 
these experiments have low copy numbers and show significantly 
lower sequence divergence than the average repeat. Murray and 
Thompson have thus been able to gain fairly specific information about 
the DNA sequences close to those regions actually functioning as 
genes in the sense of coding for proteins, though the DNA constituting 
the genes in peas is probably less than 1% of the total! 

Murray and Thompson also showed that when coding sequences are 
being expressed, they are in a conformation with respect to the associ- 
ated protein complex. This renders the coding sequences more suscep- 
tible to degradation by certain DNA-hydrolyzing enzymes. While com- 
pletely unexpected on the basis of studies with yeast, this observation 
confirms for plants what is already fairly well documented for animal 
systems, and it demonstrates for the first time that such an altered 
conformation may persist in a truly quiescent tissue. (The source of 
material was unhydrated wheat germ.) 

Marylee Everett, Richard Jorgensen, and Thompson have used clon- 
ing techniques to follow the appearance and increase in amount of two 


messenger RNAs during the greening of pea buds. These two species 
appear gradually over a period of about 72 hours in the light, after an 
initial lag of over 12 hours. Prior treatment with a small amount of red 
light eliminated this lag, so the level of the two RNAs in plants so 
treated is as high after 24 hours in the light as in controls after 72 
hours. Since the red light effect is reversible by far red, the phenome- 
non is evidently under phytochrome control. The function of the pro- 
teins produced from these messengers is unknown, but it is reasonable 
to expect that they may somehow be involved in photosynthesis. 

Richard Cuellar and Thompson report on the fine structural organi- 
zation of repetitive DNA from pea. They employed three clones pro- 
duced during studies described last year; the clones contain common 
repetitive elements, but these are interspersed with completely unre- 
lated sequences. Furthermore, certain portions of the total pea DNA 
hybridize with only one of the clones, indicating that this clone con- 
tains sequences occurring in neither of the other two. Cross- 
hybridization experiments indicated that where homology exists be- 
tween clones, the regions of homology are short. Furthermore, each 
clone contains substantially more repetitive DNA than that found in 
the common elements. Sequences in genomic DNA related to those in 
the clones show a range of repetition frequencies, and both a tandem 
and interspersed arrangement. These experiments demonstrate ele- 
gant fine tuning to document for specific fragments of DNA organiza- 
tional properties until now deducible only indirectly from studies of the 
whole genome or at best large fractions of it. 

Jeffrey Palmer and Thompson have produced some detailed studies 
on the linear organization of the chloroplast genomes from several 
species. Mung bean chloroplast DNA, like that of many other plants 
studied in several different laboratories, possesses two copies of an ap- 
proximately 23,000-base segment of DNA, one of which is inverted 
with respect to the other. It turns out that both pea and broad bean 
lack this inverted repeat. The linear sequences of homologous regions 
of DNA between pea and mung bean are very different, with one se- 
quence badly scrambled with respect to the other. By contrast, the 
linear order of sequences in spinach, cucumber, and petunia— all con- 
taining the inverted repeat— is remarkably similar to that in mung 
bean, despite the much greater evolutionary distance of these plants 
from mung bean. Broad bean shows scrambling not only with respect 
to mung bean, spinach, cucumber, and petunia, but also with respect to 
pea. Evidently, the inverted repeat confers some evolutionary stability 
to the linear arrangement of genetic material in the chloroplast DNA. 
In the absence of the inverted repeat, all sorts of inversions and 
translocations have been allowed to occur. The way in which this 
stability is conferred by the inverted repeat is completely unknown. It 


is clear, however, that the chloroplast genome may provide a very 
sharp tool for the study of plant evolution. 

Photomorphogenesis . Holly Gorton has continued her efforts to un- 
ravel the complexities of phytochrome responses in light-grown plants. 
She showed previously that responses of corn to end-of-day far red 
light exhibited certain anomalies not explainable by traditional models 
of phytochrome action. This year she has extended these studies to 
several other plants, including oats, mung bean, and sunflower. The 
anomalies, which appeared in the oat studies but not in those with 
sunflower or mung bean, are evidently not ubiquitous. Nevertheless, 
they are forcing significant changes in our concept of phytochrome ac- 
tion, and they will have to be accounted for in any future models. 

Dina Mandoli has continued her studies of the responses of dark- 
grown oat seedlings to red light by investigating in detail the site(s) of 
photosensitivity for the responses of the mesocotyl below the node and 
the coleoptile above. Surprisingly, she found maximum sensitivity to 
red light not where most of the phytochrome is found, namely in the 
node itself and in the coleoptile tip, but rather just below the node (for 
the mesocotyl response) or both just below and just above the node (for 
the coleoptile response). She also discovered that there is significant 
light piping through both organs, with fairly sharp attenuation across 
the node. By quantifying this light piping, she was able to account for 
the entire dose-response curve on the basis of reactions of the two 
photoperceptive sites to light directly absorbed at the sites as well as 
that piped there from other regions of the seedling. These findings are 
of considerable importance, particularly in view of the magnitude of 
the piping and the fact that the possibility of light piping has been vir- 
tually ignored in all previous studies. 

Jonathan Walton and James Shinkle have made significant progress 
in understanding the mechanism by which the corn mesocotyl re- 
sponds to low fluences of red light. Walton had already shown that red 
light strongly reduced the amount of the glucan synthetase localized in 
the golgi in corn mesocotyls. In the present work, he and Shinkle have 
shown that the fluence response curve for reduction of the glucan syn- 
thetase activity was almost identical to that for mesocotyl growth sup- 
pression. Both effects were somewhat reversible by far red light, and 
both could be partially potentiated by far red light alone. Auxin, which 
Vanderhoef and Briggs showed some years ago would antidote the 
growth suppression caused by red light, also antidotes the decline in 
glucan synthetase activity. Hence both growth and glucan synthetase 
activity may well be modulated by a red light-regulated supply of 

Ta-Yan Leong, continuing his detailed characterization of the light- 


sensitive flavin-cytochrome b complex in corn membrane prepara- 
tions, has shown its association with a particular ATPase, quantified 
the amount of flavin, and measured the midpoint potential of the cyto- 
chrome. Since the system is almost certainly located in the plasma 
membrane, knowledge of these properties should provide considerable 
insight into the functional nature of the membrane system limiting the 
plant cell. 

Until this year, there was little evidence linking this pigment com- 
plex with photoreception for phototropism in corn, save that both 
light-induced cytochrome reduction and phototropism were inducible 
by blue light. Recently, however, Leong has made considerable pro- 
gress by using a new class of herbicides based on a diphenyl ether 
structure. He showed that these compounds (which require light for 
their herbicidal action) significantly enhance photoreduction of the 
cytochrome. They also significantly sensitize the phototropic response 
of etiolated oat seedlings! These experiments substantially strengthen 
the case that the flavin-cytochrome complex does indeed serve as the 
photoreceptor for phototropism. Further, the herbicides provide a spe- 
cific probe which may help us to elucidate the function of this light-sen- 
sitive electron transport chain in the plant plasma membrane. 

Siegrid Schoch, Gorton, and Briggs reinvestigated the dramatic 
spectral changes, first described by Kazuo Shibata in the Department 
some 25 years ago, which occur after protochlorophyll phototransfor- 
mation in the leaves of dark-grown seedlings. These changes involve a 
shift in the absorption maximum of the newly formed chlorophyll, from 
about 684 nm to about 670 nm. Schoch and her co-workers were able to 
show (1) that carotenoids, despite their extensive presence in the pro- 
plastids, were in no way involved in the shift, (2) that reduction of the 
double bonds of the long aliphatic tail of chlorophyll was likewise not 
involved, and (3) that the shift probably did not involve proteolytic 
cleavage of the apoprotein. While we are still unaware of the physical 
or chemical changes behind the shift, we can eliminate two possibilities 
and consider a third unlikely. 

Photosynthesis. The computer capabilities in the Department of 
Plant Biology have made certain kinds of spectral analysis almost 
routine for studying a wide range of problems. Jeanette Brown and 
Schoch used the RE SOL program (which resolves complex spectra into 
simple components) to gain substantial new information about the 
spectral components associated with the two photosystems involved 
in green-plant photosynthesis. From chloroplasts of wheat and pea, 
they isolated both the light-harvesting chlorophyll-protein complex 
associated with photosystem II and the photosystem I pigment com- 
plex, and studied their spectra. These spectra could readily be modeled 


with the same four components that have served for spectra of intact 
chloroplasts in the past. However, the sum of components from the 
two preparations did not match the whole chloroplast spectrum. The 
results suggest the presence of a previously undescribed component, 
absorbing at 685 nm, which was apparently lost during the extraction 

It is well known that when plants are grown under intermittent 
light— for example 15 minutes of light every three hours— they form 
very little chlorophyll b and none of the light-harvesting chlorophyll- 
protein complex associated with photosystem II. It was therefore of 
considerable interest to compare spectra from chloroplasts of such 
plants with spectra from plants grown under normal light conditions. 
Schoch and Brown analyzed such spectra with the techniques just 
mentioned. They found that even without chlorophyll b (which absorbs 
maximally in vivo at 650 nm) there was still a substantial amount of 
650 nm absorption. This absorption must therefore arise from lower 
vibrational levels of the longer wavelength forms of chlorophyll a still 
present. Intermittent light reduced the magnitude of both the 661-nm 
and 678-nm components by the same proportion, strengthening the hy- 
pothesis that both of these bands belong to the same pigment moiety, 
a form of chlorophyll a. 

It is now well known that light is required to activate several en- 
zymes involved in carbon dioxide assimilation. Satoh has extended our 
knowledge of such photoactivation processes by showing that light is 
required to activate the enzyme ferredoxin NADP-reductase, a key en- 
zyme in the transfer of electrons from photosystem I to NADP+. Thus 
photoactivation, known to be required for several of the enzymes that 
catalyze photosynthetic carbon fixation, is also required of at least one 
enzyme mediating photosynthetic electron transport as well. The reac- 
tion probably involves a reduction of the enzyme. 

Anastasios Melis and Geoffrey Harvey continued studies started 
last year on the ratio of the two photosystem reaction centers, PSII/ 
PSI, in plants grown in different spectral environments. Melis and 
Brown had previously shown, contrary to expectations and a great 
deal of dogma, that this ratio need not be unity, but could be higher or 
lower. Melis and Harvey have now shown that any environment en- 
riched for far red light (which is preferentially absorbed by photosys- 
tem I) shows a significant increase in the number of photosystem II 
reaction centers relative to those of photosystem I. Such a shift would 
help to keep the two photoreactions in balance in a far-red-enriched en- 
vironment. Indeed, plants that grow in naturally shaded environments 
receive light enriched in far red wavelengths (because of absorption of 
red light by chlorophyll in the plants overhead); as predicted, Melis and 
Harvey found enrichment in photosystem II reaction centers in such 


plants. Chloroplasts having a high ratio of photosystem II to photo- 
system I reaction centers appear relatively enriched in granal as op- 
posed to stromal membranes. Thus differentiation at the photosystem 
level in response to far red enrichment may be accompanied by mor- 
phological differentiation. 

Over the years, there has been considerable debate about the role of 
chloroplasts in the guard cells of stomata. Recent reports have indi- 
cated that guard cells are unable to fix carbon dioxide photosyn- 
thetically. Since guard cells are the valves that regulate gas exchange 
in higher plants and since they are strongly affected by light, it is ex- 
tremely important to determine the photoreactions of which they are 
capable. Eduardo Zeiger, Paul Armond, and Melis managed for the 
first time to obtain reasonable chloroplast preparations from guard 
cells without contamination by mesophyll chloroplasts: They simply 
used albino portions of the variegated leaves of Chlorophytum com- 
osum. Zeiger and colleagues were then able to show that these chloro- 
plasts have both photosystem I and photosystem II pigments; there is 
also an intact electron transport pathway between the two photosys- 
tems. Melis and Zeiger also obtained indirect evidence that guard cell 
chloroplasts can make ATP, and that C0 2 can apparently lead to ATP 
consumption. How C0 2 leads to such photosynthetic energy consump- 
tion remains a mystery, since guard cells evidently do not carry out 
conventional photosynthetic C0 2 -fixation reactions. 

One way to learn about the effects of a given stress is to find out how 
the successful plants cope with it, an approach which has been very 
fruitful in temperature studies. To obtain good examples of resistance 
to water stress, Fork and Oquist might have travelled the well-known 
road to Death Valley. They chose instead some examples nearer at 
hand: mosses and liverworts inhabiting the surfaces of trunks and 
branches in the redwood forests near Stanford, and the red alga Por- 
phyra, which lives on intertidal rocks along the seacoast. The mosses 
and liverworts become thoroughly desiccated during the long summer 
dry season, but can function again rapidly upon rehydration; the Por- 
phyra can become dehydrated with each low tide, and in the daytime, 
can be scorched by the Sun. 

Fluorescence studies carried out at 77 K indicated that all of these 
plants lose the emission from photosystem II preferentially on desicca- 
tion. As reported last year, Porphyra simply diverts energy absorbed 
by photosystem II to photosystem I, as seen by a large increase in pho- 
tosystem I fluorescence. When an alga isolated from the lichen 
Cladonia implexa was similarly studied, it revealed a small diversion of 
energy from photosystem II to I but mostly just a loss of photosystem 
II fluorescence. In the mosses and liverworts studied, loss of photosys- 
tem II fluorescence alone was observed with no evidence of energy 


diversion. In all cases, apparent damage to photosystem II was more 
severe when drying took place in the light, suggesting that photosys- 
tem II is more susceptible to photodamage than photosystem I. For 
plants that normally become desiccated in deep shade, an energy diver- 
sion mechanism to protect photosystem II from photodamage is evi- 
dently not required, and the effects of desiccation are reversible. 

Oquist and Fork also watched fluorescence changes at room temper- 
ature in drying Porphyra, since Porphyra, unlike most species, shows 
substantial photosystem I emission at room temperature. The room- 
temperature studies amply support the conclusions from low-tempera- 
ture investigations: Upon drying, Porphyra shows a dramatic increase 
in energy transfer from photosystem II to I, and damage to photosys- 
tem II is significantly less if drying is carried out in the dark. 

With a supply of Porphyra on hand, Fork, Oquist, and George Hoch 
used fluorescence techniques to provide careful documentation that 
the long-wavelength fluorescence band from this alga, seen at room 
temperature, really does arise from photosystem I. If so, such docu- 
mentation is important, since one can then use fluorescence techniques 
and physiological temperatures to study a host of reactions associated 
with photosystem I instead of descending to 77 K. 

The Ultimate Biological Challenge 

Often characterized as the ultimate biological challenge and of 
key interest to virtually all scientists is the way in which nerve 
cells are specifically linked together to form the neural net- 
works that underlie the phenomena of perception, memory, 
and learning. 

James D. Watson 
Director's Report, 
Cold Spring Harbor Laboratory 

Cell communication— among nerve cells, between nerve cells and 
other end organs, and indeed between cells of all the tissues of the 
body— lies at the very heart of differentiation and morphogenesis. The 
work of several groups at the Department of Embryology focuses in 
one way or another at this key matter, whose ramifications extend 
widely over biological and medical science. 

Many nerves are ensheathed by glial cells, including Schwann cells 
whose function remains enigmatic. Kenneth Muller and his colleagues 
have been exploring the role of glial cells in the experimental organism, 
the medicinal leech. One theory holds that the glial-cell envelope not 
only guides the axon to the tissue it will innervate, but also may be re- 
quired for neural connections. Muller and his colleagues are able to 


observe the regeneration of a severed leech neuron as it retraces its 
path to its original target tissue, where it then makes an electrical con- 
nection. By selectively destroying the glial-cell sheath, Muller et al. 
have determined that accurate regeneration of the neuron can proceed 
in the absence of glia. 

Muller's group is not alone in the Department in their fascination 
with cell membranes and cell communication. The preparation of mono- 
clonal antibodies directed against surface molecules of muscle fibers 
provides Douglas Fambrough and his colleagues with a wealth of 
probes enabling them to detect a variety of determinants in muscle 
membranes. Antibodies against the enzyme acetylcholinesterase show 
that a patient with "myasthenia syndrome with esterase deficiency" 
has no detectable enzyme at neuromuscular junctions. This is the sec- 
ond neuromuscular disorder whose molecular basis has been elucidated 
by Fambrough and his associates. A few years ago, Fambrough and 
Daniel Drachman showed that patients with myasthenia gravis have 
decreased amounts of acetylcholine receptors at neuromuscular junc- 

A major goal of Fambrough 's group has been to explain the interac- 
tion that occurs when nerve innervates muscle. Receptors increase at 
the neuromuscular junction. An important constituent is the basal 
lamina. By using monoclonal antibodies to components of the lamina, 
Fambrough et al. have observed that the basal lamina increases when 
muscle is innervated. Newly formed receptor aggregates may be asso- 
ciated with newly synthesized basal lamina. 

Richard Pagano and his associates are devising methods to study 
the composition, metabolism, and organization of lipids in cell mem- 
branes. Their goals resemble those of Fambrough, whose emphasis is 
on proteins. The methods needed for lipid analysis, however, are very 
different from those used for proteins. Lipids are labeled isotopically or 
with fluorescent groups and are introduced into membranes by fusion. 
This method of tagging a molecule avoids labeling of the cell pool and 
introduces the molecule preformed into its metabolic pathway. Pagano 
is especially interested in determining the fate of such molecules in 
cells. He and his colleagues are now defining the variables that influ- 
ence the behavior of lipids in membranes. 

Genetic analysis promises to become increasingly powerful in study- 
ing cell communication. The motile sperm of the nematode Caenorhab- 
ditis elegans can be induced to extend their pseudopods in vitro. 
Samuel Ward has shown that this morphological change occurs by re- 
arrangement of preexisting macromolecules rather than by synthesis 
of new components. Extensive surface rearrangements accompany 
this change. Biochemical and cytological analyses of morphological 
changes alone are useful, but they are more informative when they are 


applied to sperm-specific mutants. Ward and his colleagues estimate 
that there are about 30 total genes in C. elegans that specifically affect 
sperm fertility. They have already identified mutants in 16 of them. 

Thus far we have concentrated on communication between cells. No 
less vital to our understanding of the mechanisms of development is 
the communication between nucleus and cytoplasm within a cell that is 
reflected in the regulation of gene action, which remains a prime focus 
within the Department. 

During the past 15 years a controversy has raged as to the mecha- 
nism whereby an organism can produce a seemingly unlimited variety 
of antibody molecules. Recombinant DNA and monoclonal antibody 
methods have revealed that there is not just one mechanism, but sev- 
eral. Patricia Gearhart presents convincing evidence for one of these— 
the somatic diversification of antibody genes. The first response to an 
antigen causes an immune cell to express one of a small number of 
germline heavy-chain genes (encoding a protein termed IgM). The cell 
then switches to production of proteins encoded by other heavy-chain 
genes (termed IgG or IgA). This "heavy-chain switch" not only in- 
volves a second genetic transposition but also appears to result in 
mutations within that region of the antibody gene encoding the anti- 
gen-binding part of the molecule. 

It is now some 25 years since Barbara McClintock discovered that 
genetic elements in maize can transpose in the genome, affecting genes 
they move near or within. With the advent of modern molecular ge- 
netics it has become possible to study these genetic events in great de- 
tail and to isolate the movable elements. The importance of transpos- 
able elements in eukaryotic biology becomes more evident daily. 
Transpositions appear to be responsible for the majority of spontane- 
ous mutations. Viruses that integrate into eukaryotic genomes are 
analogous in structure and behavior to these elements. Directed trans- 
positions such as those that occur in the antibody genes are now well 

There are two reasons why these movable genetic elements intrigue a 
developmental biologist. First, they cause mutations and thus provide 
a rich source of alleles with altered gene function. Second, their special 
ability to move from one genetic location to another invites a detailed 
explanation in molecular terms, and a solution might itself suggest a 
powerful tool for directing genetic change. 

Nina Fedoroff and her colleagues have concentrated on four genes in 
maize known to have mutant alleles carrying transposable elements. 
Two of these (Bronze and C2) encode enzymes on the pathway of pig- 
ment biosynthesis, while the other two (Shrunken and Waxy) encode 
enzymes of carbohydrate synthesis. Fedoroff et al. seek to describe the 


changes in the gene caused by these genetic elements. Such an analysis 
requires ultimately the isolation and characterization of wild-type and 
mutant genes. Three mutations that affect the Shrunken gene have 
been partly characterized as chromosome rearrangements. One of 
these may be an insertion within an intervening sequence of the gene. 

Transposable elements in fruit flies (Drosophila) have a different 
history from those in maize. Moderately repetitive DNA sequences 
dispersed about the genome of all plants and animals were discovered, 
originally by Roy Britten and his colleagues at the Carnegie Depart- 
ment of Terrestrial Magnetism in 1968. DNA cloning made possible 
the discovery that even closely related flies had these moderately re- 
petitive DNAs at different locations in their genomes. Gerald Rubin 
and his colleagues at the Department of Embryology, who had been in- 
volved in these discoveries, then sought to find a gene or gene complex 
that was altered by a transposable element. They now describe such a 
system, one which uses one of the best known genetic complexes— the 
white locus. Certain mutants affecting this locus were predicted and 
then shown by Rubin and his associates to be due to the insertion of 
transposable elements. An especially powerful tool for the study of 
transposable elements is a phenomenon in Drosophila called "hybrid 
dysgenesis." This appears to be caused by crossing one strain contain- 
ing many copies of a transposable element in its genome with another 
that does not. 

As part of their specialized expression, a group of genes encoding 
eggshell proteins in Drosophila are amplified just before they are ex- 
pressed. Powerful methods of molecular genetics have enabled Allan 
Spradling and his colleagues to map the regions along the genome in- 
volved in this amplification, locate the genes, and characterize the 
RNA transcripts from these genes. Using cloned fragments of this 
region, they have found previously undetected genes that appear to be 
part of the genetic control system. Mutants that rearrange regions 
within the amplified portion alter the amplification process. Spradling 
and associates hope eventually to map the origin of this DNA amplifi- 

Steven McKnight has carried out the first truly systematic analysis 
of DNA control signals in and around a protein-encoding gene. By in- 
jecting mutants of the herpes simplex thymidine kinase gene into 
Xenopus oocyte nuclei, he has delimited a region required for efficient 
transcription of the gene, between 40 and 100 nucleotides upstream 
from the start site of the gene. This region is capable of directing 
transcription initiation at a fixed distance from it even when the gene 
and its immediate flanking sequence have been deleted and replaced 
with plasmid DNA. It is clear from these and other studies that ac- 
curate initiation of protein-encoding genes, i.e., genes transcribed by 


RNA polymerase form II, require signals in the flanking region up- 
stream from the start site. This contrasts with the internal control 
region characteristic of genes transcribed by RNA polymerase form 
III, such as the 5S RNA genes. 

Attention in Donald Brown's laboratory focuses on the developmen- 
tal control of the dual 5S RNA gene system in Xenopus. Daniel 
Bogenhagen and Michael Wormington have shown that transcription 
complexes formed between" a 5S RNA gene and cellular extracts in 
vitro behave very much like chromatin isolated from nuclei. Transcrip- 
tion complexes formed both in vitro and in chromatin are exceedingly 
stable. It is even possible to show that an inactive transcription com- 
plex is stable. This stability permits a gene to be transcribed many 
times without dissociating from the factors that regulate it, and it pro- 
vides a model for the maintenance of the differential state. 


Carnegie Trustee Juan Terry Trippe and former Trustee Omar Nel- 
son Bradley died within four days of one another, in early April, in New 
York City. Several months later, the Institution lost another faithful 
supporter with the death of Trustee Emeritus James Norman White. 

Juan Trippe, air pioneer, visionary, and business leader, served the 
Institution since 1944. His was the longest trusteeship in Carnegie his- 
tory— 37 years. For many of those years, he was a member of the 
Auditing Committee, and for four years— from 1975 to 1979— its chair- 
man. Trippe retired from Pan American World Airways, the company 
he founded, in 1968. But he continued to attend the Carnegie Institu- 
tion Annual Meetings with near-perfect regularity. 

General Omar Bradley's trusteeship on the Carnegie Board lasted 
for 19 years, from 1949 to 1968. As chairman of the Retirement Benefit 
Committee (forerunner of the present Employee Benefits Committee), 
he oversaw major changes to the retirement system, entailing sizable 
increases in the Institution's annual contributions. Throughout his af- 
filiation with Carnegie, Bradley was always, in the memory of one of 
the Institution's officers, "a beautiful man to work with." 

James White, a former senior partner in the Boston investment firm 
of Scudder, Stevens and Clark, died on September 9, 1981. White 
served the Institution with great distinction from 1956 to 1979. He 
was chairman of the Board for six years, and for twelve years he served 
on the Executive Committee. White was a dedicated and loyal friend 
whose wise counsel will be sorely missed. 

George Washington Corner was director of the Department of Em- 


bryology for 15 years, from 1940 to 1955. On September 28, at the 
home of his son, George Washington Corner, Jr., he died. He was 91 
years old. In his chosen field of embryology, Corner was a giant. He de- 
veloped some of the basic concepts underlying what we know today of 
the human reproductive system. He was the co-discoverer of the fe- 
male hormone progesterone; eventually he worked out a theory of the 
hormonal basis of the menstrual cycle. A dedicated humanist, Corner 
devoted most of his remaining years to the American Philosophical 

Former Geophysical Laboratory Staff Member Emmanuel G. Zies 
died at the age of 97 in Washington, D.C., on April 26. Zies, who retired 
in 1949, worked at the Laboratory for 36 years. Described by Leason 
H. Adams, former director of the Laboratory, as "perhaps the most 
eminent living volcanologist, " Zies was a tireless investigator; his 
search for chemical information about gasses, incrustations, and other 
volcanic products took him on many trips throughout the world. 

Gregory Breit, former Staff Member (1924-1929) and Staff Associ- 
ate (1935-1937) of the Department of Terrestrial Magnetism, died in 
Salem, Oregon, on September 11, 1981. Breit's work at the DTM, to- 
gether with that of Merle Tuve, Lawrence Hafstad, and Odd Dahl, 
made possible the development of high-voltage vacuum tubes for use 
in atomic physics research. 

Sterling Hendricks, a longtime friend of the Institution and a former 
fellow at the Geophysical Laboratory, died in January at the age of 78. 
Throughout his long career as a scientist with the U.S. Department of 
Agriculture, Hendricks maintained close ties with the Carnegie Insti- 
tution. He faithfully attended the annual lectures and served for many 
years as advisor to the Departments of Embryology and Plant 

Photographer William C. Miller, who worked at the Mount Wilson 
and Palomar Observatories for 26 years (1949-1975), died on October 
8, 1981, at age 71. Miller's pioneering work in astronomical color 
photography yielded some of the first color pictures of the heavens, 
many of which are still in popular and scientific demand. 

Dorothy Rowland Swift joined the Carnegie editorial staff in 1937 
and became its chief editor shortly thereafter. For 20 years, until her 
retirement in 1957, she gave the publications of the Institution her 
careful attention. Former Carnegie president Caryl Haskins once de- 
scribed her as a "gifted editor, whose contributions to lucidity and ex- 
act expression won the respect and admiration of the staff." She died 
on April 23. 

Charles K. Ksanda served the Institution as laboratory technician 
and instrument maker for 37 years. He started in 1914 at the Geophys- 


ical Laboratory. In 1940, he was transferred to DTM, where he became 
an indispensable part of the Department's cyclotron work. In Septem- 
ber 1980, he died, at the age of 87, at his home in Los Gatos, California. 

Alfred H. Olmstead, who worked at the Hale Observatories as assis- 
tant superintendent for buildings and grounds for five years before 
retiring in 1971 at age 64, died on September 8, 1980. He served at the 
Observatories for a total of 22 years. 

Edward Gausden, who was a custodian with the Department of 
Plant Biology for seven years (1973-1980), died in Canada on April 17, 

Retired DTM instrument machinist Robert Hoffmaster passed away 
on October 2, 1981. He worked at DTM from 1963 to 1971. 

. . . AND GAINS 

John D. Macomber was elected to the Institution's Board of 
Trustees at Carnegie's Annual Meeting in May. Macomber has been 
Chairman of the Board since 1980 and Chief Executive Officer since 
1977 of the Celanese Corporation, a New York-based company that 
manufactures chemicals, fibers, and plastics. Before joining Celanese 
in 1973, he was a managing director and a member of the managing 
committee of McKinsey and Co., an international management con- 
sulting firm. 

George W. Preston, III, was appointed director of the Mount Wilson 
and Las Campanas Observatories effective July 1, 1981. Preston has 
been with the Observatories since 1968, when he was appointed Staff 
Member. From 1976 to 1980, he served as assistant director for Mount 
Wilson, and, since 1980, as acting director of the Observatories. 

Four new scientific appointments for the coming year at the Depart- 
ment of Terrestrial Magnetism reflect the Department's diverse re- 
search interests. Chemist Richard Carlson came to DTM as a postdoc- 
toral fellow from the Scripps Institution of Oceanography, La Jolla, 
California. As a Staff Member, he plans to continue his investigations 
into the nature and chronology of chemical differentiation of the 
Earth's interior using isotope and trace element studies. Staff Member 
Francois Schweizer is an astronomer from the Cerro Tololo Inter- 
American Observatory, La Serena, Chile. His interests center around 
the photometric and dynamic properties of galaxies, especially radio 
galaxies and merging galaxies. Paul Silver, from the Scripps Institution 
of Oceanography, will join the Department next year as Staff Member. 
He is interested in the use of long-period seismic waves to study lateral 
variations in the Earth's structure as well as the kinematics of the 
earthquake process. Alan Boss, who was appointed as Staff Associate, 


is a physicist from the University of California, Santa Barbara; he 
served most recently as a postdoctoral fellow in NASA's Space Divi- 
sion, Ames Research Center, California. He will study the formation of 
stars similar to the Sun in mass, and the connection between star and 
solar system formation. 

Robert Hazen, who has been with the Geophysical Laboratory since 
October 1976, first as a Research Associate and then as a Temporary 
Staff Member, has been appointed Staff Member. Hazen is concerned 
with the effects of temperature and pressure on crystal structure and 
physical properties, with the goal of better understanding the nature of 
materials deep within the Earth. An amateur historian, he and his wife 
recently compiled the 75th Anniversary Index of the Geophysical Lab- 
oratory Reports. 

The following honors were awarded to individuals during the year. 

Barbara McClintock, Distinguished Service Member of the Carnegie 
Institution, won three major prizes for her contributions to our 
understanding of chromosome structure and function and for her iden- 
tification of transposable elements. She was selected in September 
1981 as a recipient of the 1982 Wolf Prize in Medicine, to be awarded in 
March 1982 by the President of Israel at the Knesset, Israel's parlia- 
ment. In November she received the Albert Lasker Basic Medical 
Research Award and was selected as the John D. and Catherine T. 
Mac Arthur Foundation's first Prize Fellow Laureate, a prize which 
establishes an annual income of $60,000 for the rest of her life. 

In recognition of his outstanding contributions to the field of molec- 
ular developmental biology, Department of Embryology director 
Donald Brown was the co-recipient of the International Society of De- 
velopmental Biologists' first Ross G. Harrison Prize in Developmental 
Biology, awarded in Switzerland in late August 1981. In April 1981 he 
was elected to membership in the American Philosophical Society. 

George W. Wetherill, director of the Department of Terrestrial Mag- 
netism, was elected to the position of Vice-President and President- 
Elect for 1981-1983 of the Meteoritical Society at the Society's 43rd 
annual meeting, September 2-6, 1980, in San Diego. He received the 
Frederick C. Leonard Memorial Medal at the Society's 44th Annual 
Meeting in Bern, Switzerland, in August 1981. The medal is given each 
year in recognition of innovative research in meteoritics and allied 
sciences resulting in contributions to our knowledge of the origin and 
history of the solar system. 

Vera Rubin, Staff Member of DTM, was elected to the National 
Academy of Sciences in Washington on April 28, 1981. She was named 
a Chancellor's Distinguished Professor at the University of California, 
Berkeley, for the winter quarter of the 1980-1981 academic year. 


Mount Wilson Observatory Staff Member Arthur H. Vaughan deliv- 
ered the 1981 James Arthur Lecture of the Smithsonian Institution. 
He gave the lecture in May at the Harvard-Smithsonian Center for 
Astrophysics, Cambridge, Massachusetts. 

Peter M. Bell, Staff Member of the Geophysical Laboratory, was 
awarded a fellowship from the Guggenheim Memorial Foundation to 
construct geophysical and geochemical models of the Earth for inclu- 
sion in a proposed book. 

Board Chairman William R. Hewlett was elected to the American 
Philosophical Society in April 1981. 

Former Department of Embryology Staff Member Igor Dawid was 
elected to the National Academy of Sciences in April 1981. 

Tatiana Proskouriakoff, former Staff Member of the Institution's 
Department of Archaeology, was inducted into the American Philosoph- 
ical Society in April. 

In recognition of his productive research career, Roy J. Britten, Staff 
Member in Special Subject Area, was named a Distinguished Carnegie 
Senior Research Associate in Biology by the California Institute of 

In March 1981, Robert Kreidler, Vice President of the Carnegie In- 
stitution, was elected to the Board of Trustees of Barnard College. 

James Ebert was elected Vice President of the National Academy of 
Sciences in January 1981. He will serve a four-year term. 

An honor reflecting the contributions of many persons has been be- 
stowed this year on the Mount Wilson Observatory. In a formal cere- 
mony at Mount Wilson on June 20, 1981, the American Society of Me- 
chanical Engineers (ASME) dedicated the 100-inch Hooker Telescope 
as an International Historic Mechanical Engineering Landmark. In his 
remarks during the presentation, Dr. Robert Gaither, President of 
ASME, stressed the role in science of those "men and women who take 
the dreams and the fantasies of a society and convert them into operat- 
ing hardware"— the mechanical engineers. 

The dedication illuminates and reinforces George Ellery Hale's state- 
ment quoted earlier in this essay, that a "single improvement in instru- 
ments may render years of work with inferior apparatus unnecessary." 
For three decades after its completion in 1918, the innovative Hooker 
Telescope remained the world's largest. Its use made possible perhaps 
the most significant research in modern cosmology, including the dis- 
covery of the expanding universe by Milton Humason and Edwin 
Hubble. Among its outstanding features are its mirror support and the 
use of mercury flotation to reduce friction. Today the telescope, imagi- 
natively fitted with new instrumentation, continues to play a vital role 
in astronomical research. 

James D. Ebert 




Stanford, California 

Winslow R. Briggs 

Director Emeritus 
C. Stacy French 

William M. Hiesey 

Staff Members 
Joseph A. Berry 
Olle Bjorkman 
Jeanette S. Brown 
David C. Fork 
Malcolm A. Nobs 
William F. Thompson 

Research Associate 
Michael G. Murray 

Senior Fellows 
Mary lee Everett 
George E. Hoch 
Dov Roller 
Jacob Levitt 
Gunnar Oquist 
Kazuhiko Satoh 
Siegrid Schoch 

Geoffrey W. Harvey 
Richard A. Jorgensen 
Ta-Yan Leong 
Stephen B. Powles 
Anastasios Melis 

Richard E. Cuellar 
Holly L. Gorton 
Dina F. Mandoli 
Jeffrey D. Palmer 
Terri L. Reichert 
John Schaer 

Jeffrey R. Seemann 
James Shinkle 
William E. Williams 

Baltimore, Maryland 

Donald D. Brown 

Staff Members 
Douglas M. Fambrough 
Nina V. Fedoroff 
Kenneth J. Muller 
Richard E. Pagano 
Gerald M. Rubin 
Allan Spradling 
Samuel Ward 

Staff Associates 
Patricia Gearhart 
Steven L. Mc Knight 
Richard Rotundo 

Research Associates and Fellows 
M. John Anderson 
Ellen Bayne 
Daniel Bogenhagen 
Deborah Chaleff 
Matthias Chiquet 
Mary Collins 
Ellen Elliott 
Mark Emmerling 
Dick Hoekstra 
Ian Jackson 
Roger Karess 
Robert Levis 
Adrian Mason 
Sheila McCormick 
Wylie Nichols 
Kevin O'Hare 
Hugh Pelham 
Ronald Peterson 
Thomas Roberts 
Alan Schroit 



Mavis Shure 
Douglas Struck 
Adina Student 
Barbara Wakimoto 
Eric Wakshull 
Susan Wessler 
Michael Wormington 

Visiting Investigators 
J. G. Gall 
K. Longmuir 
L. Mets 

Christopher Austin 
Dan Burke 
John M. Gardner 
Elizabeth Gavis 
Elizabeth Liebson 
Benjamin Nathans 
Suki Parks 
Mark Schlissel 
Jennifer Schwartz 

Washington, D.C. 


Hatten S. Yoder, Jr. 

Gordon L. Davis 
Elburt F. Osborn 
Emanuel G. Zies 1 

Staff Members 
Peter M. Bell 
Francis R. Boyd, Jr. 
Felix Chayes 
Marilyn L. F. Estep 
Larry W. Finger 
John D. Frantz 
P. Edgar Hare 
Robert M. Hazen 
Thomas C. Hoering 
T. Neil Irvine 
Ikuo Kushiro 
Ho-Kwang Mao 
Tsutomu Murase 
Bjorn O. Mysen 

'Died, April 26, 1981 

Douglas Rumble III 
Friedrich A. Seifert 
David Virgo 

Timothy M. Benjamin 
Nabil Z. Boctor 
Paul A. Danckwerth 
Michael H. Engel 
Steven W. Lonker 
Eiichi Takahashi 

Predoctoral Fellows and Students 
Julia C. Corrado 
Donald L. Elthon 
Andrew P. Gize 
John M. Hughes 
John Lambie 
Antonio M. G. Possolo 
Russell L. Ralph 
Charles M. Schlinger 



Washington, D.C. 


George W. Wetherill 

Distinguished Service Member 
Merle A. Tuve 


Scott E. Forbush 

Staff Members 
L. Thomas Aldrich 
Louis Brown 
W. Kent Ford, Jr. 
Albrecht W. Hofmann 
David E. James 
Typhoon Lee 
Alan T. Linde 
R. Sundar Raj an 
Vera C. Rubin 
I. Selwyn Sacks 
Fouad Tera 
Norbert Thonnard 

Research Associates 
Mark Harrison 
Leonidas Ocola 



Senior Fellow 
Arvind S. Tamhane 

Richard W. Carlson 
Bruce W. Carney 
Lina Maria Echeverria 
Esther M. Hu 
Emi Ito 

Brajesh K. Kothari 
Kiyoji Shiono 
Robert J. Stern 
Bradley C. Whitmore 

Predoctoral Fellows and Students 
Thomas M. Boyd 
Michael So Chyi 
Diana Diez-de-Medina 
Morgan I. Granger 
William K. Hart 
James M. Horan 
Yukiko Maeda 
Nancy J. Sklower 

Visiting Investigators 
Douglas G. Mose 
Gerard R. Poupeau 
David Schwartzman 
J. Arthur Snoke 
Ragnar Stefansson 
Richard T. Williams 
Xinhua Zhou 

Pasadena, California 

George W. Preston 

Horace W. Babcock 
Olin C. Wilson 

Staff Members 
Halton C. Arp 
Robert F. Howard 
Jerome Kristian 
S. Eric Persson 

Allan R. Sandage 
Leonard Searle 
Stephen A. Shectman 
Arthur H. Vaughan 

Post-Retirement Studies 
Henrietta H. Swope 2 

Staff Associates 
Robert J. Brucato 
Jean J. Lorre 

Senior Fellows 
Alan M. Dressier 
Thomas R. Geballe 

David H. Bruning 
Douglas K. Duncan 
Debra M. Elmegreen 
Howard B. French 
Robert C. Kennicutt 
Barry J. LaBonte 
Armando Manduca 
Horace A. Smith 
Peter B. Stetson 
Steven G. Wallenhorst 

Carnegie-Chile Fellow 
Guido Garay 

Student Observers 
Graham Berriman 
Kirk D. Borne 
Alexei Filippenko 
Richard Gomer 
Keith Home 
Matthew Malkan 
Stefan Mocknacki 
Daniel Nadeau 
Beth Nordholt 
Jeffrey R. Pier 
Richard Pogge 
Carolyn Porco 
R. Michael Rich 
Raghvendra Sahai 
Kris ten Sellgren 

2 Died, November 24, 1980 

Reports of Departments and 
Special Studies 

Department of Plant Biology 

Department of Embryology 

Developmental Biology Research Group 

Geophysical Laboratory 

Department of Terrestrial Magnetism 

Mount Wilson and Las Campanas Observatories 

Program in Science Policy 

Department of Plant Biology 

Stanford, California 

Winslow R. Briggs 



Introduction (Briggs) 7 

Spectral analyses of chlorophyll- 
protein complexes (Brown and 
Schoch) 15 

Comparative spectroscopy of chloro- 
phyll a in daylight- and intermit- 
tent-light-grown plants (Schoch 
and Brown) 16 

Fluorescence induction and photoacti- 
vation of ferredoxin-NADP + re- 
ductase in Bryopsis chloroplasts 
(Kazuhiko Satoh) 20 

Regulation of photosystem stoich- 
iometry and relation to chloroplast 
ultrastructure (Melis and Harvey) 25 

Fluorescence properties of guard cell 
chloroplasts: evidence for linear 
electron transport and light-har- 
vesting pigments of photosystems 
I and II (Zeiger, Armond, and 
Melis) 29 

Fluorescence properties of guard cell 
chloroplasts: evidence for CO2 
modulation of photophosphoryla- 
tion (Melis and Zeiger) 31 

Effects of desiccation on the excitation 
energy distribution in the red alga 
Porphyra perforata, the liverwort 
Porella navicularis, and the 
isolated lichen green alga Tre- 
bouxia pyriformis (Oquist and 
Fork) 34 

Desiccation effects on transfer of ex- 
citation energy between the two 
photosystems of photosynthesis at 
physiological temperatures in the 
red alga Porphyra perforata (Fork 
and Oquist) 39 

Room-temperature photosystem I 
fluorescence emission in the red 
alga Porphyra perforata (Fork, 
Oquist, and Hoch) 


Reoxidation of the primary electron 
acceptor of photosystem II at low 
temperature (Fork and Oquist) .... 45 

Solubilization and spectral charac- 
teristics of chlorophyll-protein 
complexes isolated from the ther- 
mophilic blue-green alga Synecho- 
coccus lividus (Oquist, Fork, 
Schoch, Malmberg) 47 

A fluorescence decline as an indicator 
of photoinhibition in intact Bry- 
opsis chloroplasts under anaerobic 
conditions (Satoh and Fork) 50 

Photoinhibition in bean: a fluorescence 
analysis (Fork, Oquist, and 
Powles) 52 

Interaction between high irradiance 
and water stress on photosyn- 
thetic reactions in Nerium oleander 
(Bjorkman, Powles, Fork, and 
Oquist) 57 

Leaf movement in the shade species 
Oxalis oregana. I. Response to 
light level and light quality (Bjork- 
man and Powles) 59 

Leaf movement in the shade species 
Oxalis oregana. II. Role in pro- 
tection against injury by intense 
light (Powles and Bjorkman) 63 

The relationship between photosyn- 
thetic performance and the levels 
and kinetic properties of RuBP 
carboxylase-oxygenase from des- 
ert winter annuals (Seemann, 
Tepperman, and Berry) 67 

Solar tracking (phototropism) in leaves 
of Lavatera cretica and Malva 
parviflora (Roller) 72 

Studies on mRNA-coding DNA in 

higher plants (Murray and Thomp- 
son) 76 

Phytochrome control of transcript 
abundance in developing pea 
leaves (Everett, Jorgensen, and 
Thompson) 79 

Complex organization of repetitive 
DNA families, as analyzed with 
cloned DNA fragments (Cuellar 
and Thompson) 81 

Evolutionary stability of the higher- 
plant chloroplast genome (Palmer 
and Thompson) 82 

A survey of the reciprocity relation- 
ships for responses to end-of-day 
irradiations in four plants (Gorton 
and Briggs) ..." 85 

Localization of the region(s) of photo- 
sensitivity in the etiolated Auena 
seedling (Mandoli and Briggs) .... 88 

Red light inhibition of growth and 
golgi-localized glucan synthetase 
activity in the maize mesocotyl 
(Walton, Shinkle, and Briggs) ... 90 

Further characterization of a blue- 
light-sensitive cytochrome-flavin 

complex from corn coleoptile mem- 
branes (Leong and Briggs) 93 

Effect of diphenyl ethers on the blue- 
light-induced absorbance change, 
phototropism, and geotropism in 
etiolated corn and oat seedlings 
(Leong and Briggs) . . . 94 

Investigations of the Shibata shift in 
etiolated primary leaves of corn 
and oat seedlings (Schoch, Gorton, 
and Briggs) 96 

Bibliography 99 

Speeches 100 

Personnel 104 


During the past decade, a significant 
portion of the Department of Plant 
Biology's efforts has been devoted to 
learning how desert species tolerate 
the environmental extremes to which 
they are exposed. Until relatively re- 
cently, these studies have concentrated 
on high-temperature stress. Thus Bjork- 
man, Berry, and their associates have 
learned a great deal about the kinds of 
photosynthetic carbon pathways in 
plants adapted to high temperature, the 
components in a membrane that enable 
it to remain functional at high tempera- 
tures, and finally how those plants that 
can grow at wide ranges of temperature 
modify their biochemistry in response to 
changing temperature— a process known 
as acclimation. 

The past year has brought a signifi- 
cant change in emphasis. Of the reports 
that follow, only one deals with acclima- 
tion and only two with plants from arid 
regions, of which neither addresses prob- 
lems of temperature stress. However, 
there are three articles dealing with 
water stress and five with photoinhibi- 
tion (damage to photosynthetic capacity 
occurring when more light is absorbed 
by a leaf than can be used by normal 
photosynthetic reactions). Photoinhibi- 
tion can be induced by any factor re- 
stricting utilization of the reducing 
power generated by light-driven electron 
transport. Low temperature, limitation 
in C0 2 availability, and water stress can 
in combination with high light intensity 
all lead to photoinhibition and, as will 
become clear below, all ultimately pro- 
duce similar kinds of damage. Since in 
arid regions any or all of these factors 
may obtain, a knowledge of their specific 
effects is crucial to complement our al- 
ready extensive knowledge of high-tem- 
perature effects. Such considerations 
make the shift in research direction both 
logical and timely. 

In singling out these studies on the ef- 
fects of stress, there is no intention to 
slight other areas of research in the De- 
partment. Occasionally, however, there 
is a turning point in the Department's 
research or the emergence of a new 
theme which merits special mention. 
This year, the stress research provides 
such a case. 


Our present computer capabilities have 
made certain kinds of spectral analysis 
almost routine for studying a wide range 
of problems. Brown and Schoch used the 
RE SOL program (which resolves com- 
plex spectra into simple components) to 
gain substantial new information about 
the spectral components associated with 
the two photosystems involved in green 
plant photosynthesis. From chloroplasts 
of wheat and pea, they isolated both the 
light-harvesting chlorophyll-protein com- 
plex associated with photosystem II and 
the photosystem I pigment complex, 
and studied their spectra. These spectra 
could readily be modeled with the same 
four components that have served for 
spectra of intact chloroplasts in the past. 
However, the sum of components from 
the two preparations did not match the 
whole chloroplast spectrum. The results 
suggest the presence of a previously 
undescribed component, absorbing at 
685 nm, which was apparently lost dur- 
ing the extraction procedure. 

It is well known that when plants are 
grown under intermittent light— for 
example, 15 minutes of light every 
three hours— they form very little chlo- 
rophyll b and none of the light-harvest- 
ing chlorophyll-protein complex asso- 
ciated with photosystem II. It was 
therefore of considerable interest to 
compare spectra from chloroplasts of 



such plants with spectra from plants 
grown under normal light conditions. 
Schoch and Brown analyzed such spec- 
tra with the techniques just mentioned. 
They found that even without chloro- 
phyll b (which absorbs maximally in 
vivo at 650 nm) there was still a sub- 
stantial amount of 650-nm absorption. 
This absorption must therefore arise 
from lower vibrational levels of the 
longer wavelength forms of chlorophyll 
a still present. Intermittent light re- 
duced the magnitude of both the 661-nm 
and 678-nm components by the same 
proportion, strengthening the hypoth- 
esis that both of these bands belong to 
the same pigment moiety, a form of 
chlorophyll a. 

Although this work does not appear 
below in a formal report, Brown has 
made considerable progress in develop- 
ing and extending techniques for isola- 
tion and characterization of a variety of 
chlorophyll-protein complexes, both 
from higher plants and from a series of 
microalgae not previously studied. 
Brown carried out this work while on 
sabbatical leave in Australia during the 
academic year 1980-1981. These new 
techniques, along with new experimen- 
tal material for the various complexes, 
should substantially enhance her ef- 
forts to characterize these complexes in 
as unaltered a form as possible and to 
learn how they interact. 

This past year there have been two 
other studies of the intrinsic properties 
of chloroplast photosynthetic ma- 
chinery. It is now well known that light 
is required to activate several enzymes 
involved in carbon dioxide assimila- 
tion. Satoh has extended our know- 
ledge of such photoactivation processes 
by showing that light is required to ac- 
tivate the enzyme ferredoxin NADP- 
reductase, a key enzyme in the transfer 
of electrons from photosystem I to 
NADP+. Thus photoactivation, known 
to be required for several of the en- 
zymes that catalyze photosynthetic 
carbon fixation, is also required of at 
least one enzyme mediating photosyn- 

thetic electron transport as well. The 
reaction probably involves a reduction 
of the enzyme. 

Melis and Harvey continued studies 
started last year on the ratio of the two 
photosystem reaction centers, PS 11/ 
PS I, in plants grown in different spec- 
tral environments. Melis and Brown 
had previously shown, contrary to ex- 
pectations and a great deal of dogma, 
that this ratio need not be unity but 
could be higher or lower. Melis and 
Harvey have now shown that any en- 
vironment enriched for far-red light 
(which is preferentially absorbed by 
photosystem I) shows a significant in- 
crease in the number of photosystem II 
reaction centers relative to those of 
photosystem I. Such a shift would help 
keep the two photoreactions in balance 
in a far-red-enriched environment. In- 
deed, plants that grow in naturally 
shaded environments receive light en- 
riched in far-red wavelengths (because 
of absorption of red light by chloro- 
phyll in the plants overhead); as pre- 
dicted, Melis and Harvey found enrich- 
ment in photosystem II reaction centers 
in such plants. Chloroplasts having a 
high ratio of photosystem II to photo- 
system I reaction centers appear rela- 
tively enriched in granal as opposed to 
stromal membranes. Thus differentia- 
tion at the photosystem level in re- 
sponse to far-red enrichment may be 
accompanied by morphological differ- 

Over the years, there has been con- 
siderable debate about the role of chlo- 
roplasts in the guard cells of stomata. 
Recent reports have indicated that 
guard cells are unable to fix carbon di- 
oxide photosynthetically. Since guard 
cells are the valves that regulate gas 
exchange in higher plants and since 
they are strongly affected by light, it is 
extremely important to determine the 
photoreactions of which they are ca- 
pable. Zeiger, Armond, and Melis man- 
aged for the first time to obtain reason- 
able chloroplast preparations from 
guard cells without contamination by 


mesophyll chloroplasts: They simply 
used albino portions of the variegated 
leaves of Chlorophytum comosum. 
Zeiger and colleagues were then able to 
show that these chloroplasts have both 
photosystem I and photosystem II 
pigments; there is also an intact elec- 
tron transport pathway between the 
two photosystems. Melis and Zeiger 
also obtained indirect evidence that 
guard cell chloroplasts can make ATP, 
and that C0 2 can apparently lead to 
ATP consumption. How C0 2 leads to 
such photosynthetic energy consump- 
tion remains a mystery, since guard 
cells evidently do not carry out conven- 
tional photosynthetic C0 2 -fixation re- 

As mentioned above, one way to learn 
about the effects of a given stress is to 
find out how the successful plants cope 
with it, an approach which has been 
very fruitful in temperature studies. To 
obtain good examples of resistance to 
water stress, Fork and Oquist might 
have traveled the well-known road to 
Death Valley. They chose instead some 
examples nearer at hand: mosses and 
liverworts inhabiting the surfaces of 
trunks and branches in the redwood 
forests near Stanford, and the red alga 
Porphyra, which lives on intertidal 
rocks along the seacoast. The mosses 
and liverworts become thoroughly des- 
iccated during the long summer dry 
season, but can function again rapidly 
upon rehydration; Porphyra can be- 
come dehydrated with each low tide, 
and in the daytime, can be simultane- 
ously scorched by the sun. 

Fluorescence studies carried out at 
77 K indicated that all of these plants 
lose the emission from photosystem II 
preferentially on desiccation. As re- 
ported last year, Porphyra simply di- 
verts energy absorbed by photosystem 
II to photosystem I, as seen by a large 
increase in photosystem I fluorescence. 
When an alga isolated from the lichen 
Cladonia implexa was similarly studied, 
it revealed a small diversion of energy 
from photosystem II to I but mostly 

just a loss of photosystem II fluores- 
cence. In the mosses and liverworts 
studied, loss of photosystem II fluo- 
rescence alone was observed with no evi- 
dence of energy diversion. In all cases, 
apparent damage to photosystem II 
was more severe when drying occurred 
in the light, suggesting that photo- 
system II is more susceptible to photo- 
damage than photosystem I. For plants 
that normally become desiccated in deep 
shade, an energy diversion mechanism to 
protect photosystem II from photodam- 
age is evidently not required, and the ef- 
fects of desiccation are reversible. 

Oquist and Fork also watched flu- 
orescence changes at room tempera- 
ture in drying Porphyra, since Por- 
phyra, unlike most species, shows sub- 
stantial photosystem I emission at 
room temperature. The room-tempera- 
ture studies amply support conclusions 
from low-temperature investigations: 
upon drying, Porphyra shows a dra- 
matic increase in energy transfer from 
photosystem II to I, and damage to 
photosystem II is significantly less if 
drying is carried out in the dark. 

With a supply of Porphyra on hand, 
Fork, Oquist, and Hoch used fluo- 
rescence techniques to provide careful 
documentation that the long-wave- 
length fluorescence band from this 
alga, seen at room temperature, really 
does arise from photosystem I. If so, 
such documentation is important, 
since one can then use fluorescence 
techniques and physiological tempera- 
tures to study a host of reactions asso- 
ciated with photosystem I instead of 
descending to 77 K. 

In a departure from their desiccation 
studies, Fork and Oquist followed the 
dark reoxidation of the primary elec- 
tron acceptor for photosystem II at 
77 K (by studying fluorescence). The 
dark reoxidation was biphasic, and 
could be resolved into two components 
with very different rate constants. 
This suggests that the photosystem II 
centers possess some heterogeneity. 
Since the light-induced reduction of the 



centers is also known to be biphasic, 
suggesting heterogeneity, the results 
of Fork and Oquist nicely complement 
the existing literature. 

In the sole study dealing with ther- 
mal acclimation, Oquist and Fork, with 
Schoch and Malmberg, followed the 
relative ease with which various photo- 
synthetic pigment-protein complexes 
could be solubilized by treating with 
detergent the membranes from the 
thermophilic alga Synechococcus livi- 
dus grown at two different tempera- 
tures. Attempted solubilization at a 
series of different temperatures re- 
vealed several effects of growth tem- 
perature. Most prominently, the photo- 
system I antenna pigment complex, 
CP-aj, was quite inaccessible to deter- 
gent below the phase transition tem- 
perature for the membrane lipids. Fork 
had previously shown that this phase 
transition temperature was higher 
when the plants were grown at the 
higher temperature. 

The next five reports deal with dif- 
ferent manifestations of the phenom- 
enon of photoinhibition. In the first of 
these studies, Satoh and Fork investi- 
gated the effects of anaerobiosis on the 
photosynthetic chloroplasts from the 
green alga Bryopsis. Chloroplasts in 
excellent condition can readily be ob- 
tained from this organism, making it a 
particularly favorable material for stud- 
ies of injury effects. It is known that ox- 
ygen can serve as a competing electron ac- 
ceptor for photosynthetically generated 
electrons. Satoh and Fork reasoned that if 
this was so, oxygen could actually act as a 
protectant when conditions leading to 
photoinhibition prevailed. This predic- 
tion was verified by the demonstration 
that even under very low light fluence 
rates, photoinhibitory damage could be 
detected if oxygen was excluded. 

Last year, Powles, Berry, and Bjork- 
man showed that at low temperatures, 
for example at 6°C, high light intensity 
led to photoinhibition in the chilling- 
sensitive bean, Phaseolus vulgaris. This 
year, Fork, Oquist, and Powles investi- 

gated the effects of such photoin- 
hibitory treatment on photosystems I 
and II. Photoinhibitory treatment 
clearly led to the accumulation of inac- 
tive photosystem II reaction centers. 
The phenomenon was detectable both 
with intact leaves (fluorescence mea- 
sured from the upper leaf surface) and 
with isolated chloroplasts. However, 
the fluorescence measurements of the 
leaves showed the effect much more 
dramatically, indicating that chloro- 
plasts in the upper surface of the leaf 
are the most damaged. The extracted 
chloroplasts, of course, represent 
chloroplasts from throughout the leaf, 
not just the upper surface. 

Physiological Ecology 

The distinction between Photosyn- 
thesis and Physiological Ecology at 
this point represents mostly a change 
in laboratories (and in experimental 
plants). In both areas, the experimental 
approach still includes fluorescence 
measurements; both groups ask, as an 
underlying question, what are the con- 
sequences of stress? 

There is even a substantial overlap of 
authorship: Bjorkman, Powles, Fork, 
and Oquist investigated the effects of 
water stress on photosynthetic per- 
formance of the shrub Nerium oleander 
in full sunlight and under shaded con- 
ditions. They found that water stress 
damaged whole-chain electron trans- 
port far more than it damaged photo- 
system I electron transport, and they 
verified that there was a strong effect 
of water stress on photosystem II. 
Fluorescence measurements indicated 
that water stress leads to photoin- 
hibitory inactivation of photosystem II 
reaction centers precisely as it does for 
bean under chilling conditions (see 

The final two papers on photoinhibi- 
tion take a rather different turn. The 
mobile laboratory also took a different 
turn, heading not for Death Valley but 



for San Mateo County Redwood Memo- 
rial Park, on the west side of the Santa 
Cruz Mountains. There, on the forest 
floor, in a deeply shaded environment, 
can be found a variety of small plants, 
among them Oxalis oregana. The 
leaves of this species consist of three 
leaflets, each attached to a vertical 
petiole by its own pulvinus. By chang- 
ing the relative turgor in different 
groups of cells, the pulvini regulate the 
position of the leaflets with respect to 
the petiole over a 90° range. Normally, 
the light intensity incident on these 
plants is about 0.5% that of full 
sunlight. However, occasional sun- 
flecks pass over them, and may persist 
for an hour or more. On arrival of the 
sunfleck, the light intensity may in- 
crease as much as 200-fold. 

Bjorkman and Powles found that 
when such a jump occurs, the leaves 
fold rapidly downward, exposing only a 
small area to the full sunlight. The 
folding response was remarkably ra- 
pid—the lag period was as brief as six 
seconds, and folding was complete 
within six minutes. Recovery after pas- 
sage of the sunfleck was slower, show- 
ing about a 10-minute lag and requir- 
ing another 35 minutes for restoration 
of the original horizontal leaflet posi- 

Light intensities of 15-30% of full 
sunlight are sufficient to trigger the 
folding response. Above the threshold 
intensities, the lag period becomes 
shorter and the angular rate of position 
change more rapid. It is the pulvini 
themselves that sense the changes in 
incident light intensity, and the effec- 
tive wavelengths are in the blue. Per- 
haps the photoreceptor is similar to 
those for solar tracking by leaves and 
for corn phototropism discussed else- 
where in this Report. 

By making gas exchange measure- 
ments with the mobile laboratory, 
Powles and Bjorkman were able to 
demonstrate convincingly that the 
folding reaction forestalled photoinhi- 
bitory damage. If the Oxalis leaves 

were constrained from folding when a 
sunfleck passed by, substantial photo- 
inhibition occurred, requiring several 
hours for recovery. Curiously, the 
folding reaction scarcely affected 
photosynthesis rates. Photosynthesis 
in these shade plants is normally sat- 
urated at very low light intensities. 
Hence, even when they are folded and 
presenting only a very small area to 
the sun, there is still sufficient light in 
the sunfleck to saturate their photo- 
synthetic machinery. 

Two other studies complete the 
year's contributions in physiological 
ecology. In the first of these, Seemann, 
Tepperman, and Berry report on con- 
tinuation of work described last year 
on the photosynthetic capacities of 
some of the annual plants from Death 
Valley. These plants by and large have 
photosynthetic capacities well above 
those of most temperate species: 
species of Camissonia show the highest 
values recorded to date. In most cases, 
the very high photosynthetic capacity 
could be accounted for by unusually 
high levels of the enzymes required for 
fixing C0 2 into organic matter, most 
notably the enzyme that initially binds 
the C0 2 for subsequent photosynthetic 
steps, ribulose-l,5-bisphosphate car- 
boxylase-oxygenase. (After years of 
struggling with difficult and uneupho- 
nious acronyms for this enzyme, workers 
have finally suggested the nickname 
"Rubisco." At least it is pronounce- 
able.) With Camissonia, however, there 
is another important difference. The 
Rubisco from Camissonia has a spe- 
cific activity significantly higher than 
that of any other species studied. Thus, 
not only is there more of it, but it is 
more efficient. The basis for this 
greater efficiency is at present un- 

The final study in this section con- 
cerns the phenomenon of sun-tracking. 
The leaves of certain species have the 
capacity to follow the sun with great 
precision, hence maximizing the 
interception of incoming solar energy. 



Roller has been studying solar track- 
ing in two such species— Lavatera 
cretica and Malva parviflora— by 
means of a tracking device which per- 
mits him to sustain a given angle for 
incident light with respect to the leaf 
surface and to study the tracking as a 
steady-state phenomenon. Leaves of 
both species can undergo reorientation 
at a high velocity for as much as an 
hour, and the rates can far exceed the 
earth's rotation (15° per hour). Roller 
has recorded rates as high as 90° per 
hour, and has found that leaves can 
continue to coast for some time in 
darkness following driving illumina- 
tion. The stimulus itself was shown 
earlier by Roller and Schwartz to be 
perceived not by the pulvinus itself, as 
is the case for the leaf-folding Oxalis 
response, but rather by the major leaf 
veins. Some influence is then transmit- 
ted down the veins to the pulvinus, 
which is located where the leaf blade is 
attached to the petiole. Turgor changes 
in the pulvinus mediate most of the 
leaf movement. There is evidence, 
however, that the petiole itself can also 
participate; functional differentiation 
of pulvinus from petiole is hence in- 
complete. The leaves can reorient 
either toward or away from the stem, 
and both processes probably involve 
active transport phenomena. Leaf 
tracking is yet another process reg- 
ulated by a blue light photoreceptor. 

Molecular Biology 

Several years ago, a Trustees' Vis- 
iting Committee referred to the tools of 
plant molecular biology as highly pro- 
mising but still in need of substantial 
sharpening. Comparison of the reports 
from the Molecular Biology group five 
years ago with those in this Report will 
indicate immediately the dramatic ex- 
tent to which sharpening has occurred. 
Murray and Thompson, continuing 
their detailed analysis of the pea's 
genome structure, have inquired as to 

the nature of the DNA sequences in the 
neighborhood of those sequences that 
actually code for messenger RNA. Sur- 
prisingly, the coding sequences are 
enriched in DNA regions where there is 
a paucity of highly repetitive DNA, 
and are depleted in regions wherein 
highly repetitive elements abound. 
Those repeats which co-isolate with 
gene sequences in these experiments 
have low copy numbers and show sig- 
nificantly lower sequence divergence 
than the average repeat. Murray and 
Thompson have thus been able to gain 
fairly specific information about the 
DNA sequences close to those regions 
actually functioning as genes in the 
sense of coding for proteins, though 
the DNA constituting the genes in 
peas is probably less than 1% of the 

Murray and Thompson also showed 
that when coding sequences are being ex- 
pressed, they are in a special conforma- 
tion with respect to the associated pro- 
tein complex. This renders the coding 
sequences more susceptible to degrada- 
tion by certain DNA-hydrolyzing en- 
zymes. While completely unexpected on 
the basis of studies with yeast, this 
observation confirms for plants what is 
already fairly well documented for ani- 
mal systems, and it demonstrates for the 
first time that such an altered conforma- 
tion may persist in a truly quiescent 
tissue. (The source of material was unhy- 
drated wheat germ.) 

Everett, Jorgensen, and Thompson 
have used cloning techniques to follow 
the appearance and increase in amount 
of two messenger RNAs during the 
greening of pea buds. These two spe- 
cies appear gradually over a period of 
about 72 hours in the light, after an ini- 
tial lag of over 12 hours. Prior treat- 
ment with a small amount of red light 
eliminated this lag, so the level of the 
two RNAs in plants so treated is as 
high after 24 hours in the light as in 
controls after 72 hours. Since the red 
light effect is reversible by far red, the 
phenomenon is evidently under phyto- 



chrome control. The function of the 
proteins produced from these mes- 
sengers is unknown, but it is reason- 
able to expect that they may somehow 
be involved in photosynthesis. 

Cuellar and Thompson report on the 
fine structural organization of repeti- 
tive DNA from pea. They employed 
three clones produced during studies 
described last year; the clones contain 
common repetitive elements, but these 
are interspersed with completely un- 
related sequences. Furthermore, cer- 
tain portions of the total pea DNA 
hybridize with only one of the clones, 
indicating that this clone contains se- 
quences occurring in neither of the 
other two. Cross-hybridization experi- 
ments indicated that where homology 
exists between clones, the regions of 
homology are short. Furthermore, each 
clone contains substantially more re- 
petitive DNA than that found in the 
common elements. Sequences in gen- 
omic DNA related to those in the 
clones show a range of repetition fre- 
quencies and both a tandem and inter- 
spersed arrangement. These experi- 
ments demonstrate elegant fine tuning 
to document for specific fragments of 
DNA organizational properties until 
now deducible only indirectly from 
studies of the whole genome or at best 
large fractions of it. 

Finally, Palmer and Thompson have 
produced some detailed studies on the 
linear organization of the chloroplast 
genomes from several species. Mung 
bean chloroplast DNA, like that of 
many other plants studied in several 
different laboratories, possesses two 
copies of an ~23,000-base segment of 
DNA, one of which is inverted with re- 
spect to the other. It turns out that 
both pea and broad bean lack this in- 
verted repeat. The linear sequences of 
homologous regions of DNA between 
pea and mung bean are very different, 
with one sequence badly scrambled 
with respect to the other. By contrast, 
the linear order of sequences in spin- 
ach, cucumber, and petunia— all con- 

taining the inverted repeat— is remark- 
ably similar to that in mung bean, 
despite the much greater evolutionary 
distance of these plants from mung 
bean. Broad bean shows scrambling 
not only with respect to mung bean, 
spinach, cucumber, and petunia, but 
also with respect to pea. Evidently, the 
inverted repeat confers some evolu- 
tionary stability to the linear arrange- 
ment of genetic material in the chloro- 
plast DNA. In the absence of the in- 
verted repeat, all sorts of inversions 
and translocations have been allowed 
to occur. The way in which this stabil- 
ity is conferred by the inverted repeat 
is completely unknown. It is clear, 
however, that the chloroplast genome 
may provide a very sharp tool for the 
study of plant evolution. 

Photomorpho gene sis 

Although our formal discussion of re- 
search on photomorphogenesis begins 
here, we have already mentioned blue 
light photoreceptors twice and phyto- 
chrome once. This seeming anomaly in- 
dicates two things: (1) the artificiality 
of the current divisions of the Depart- 
ment, and (2) more important, the kind 
of cross-fertilization that frequently oc- 
curs between divisions. The studies of 
photoinhibition provide another strik- 
ing example of this phenomenon. 

Gorton has continued her efforts to 
unravel the complexities of phyto- 
chrome responses in light-grown plants. 
She showed previously that responses 
of corn to end-of-day far-red light ex- 
hibited certain anomalies not explain- 
able by traditional models of phyto- 
chrome action. This year she extended 
these studies to several other plants, 
including oat, mung bean, and sun- 
flower. The anomalies, which appeared in 
the oat studies but not in those with 
sunflower or mung bean, are evidently 
not ubiquitous. Nevertheless, they are 
forcing significant changes in our con- 
cept of phytochrome action, and they 



will have to be accounted for in any 
future models. 

Mandoli has continued her studies of 
the responses of dark-grown oat seed- 
lings to red light by investigating in 
detail the site(s) of photosensitivity for 
the responses of the mesocotyl below 
the node and the coleoptile above. Sur- 
prisingly, she found maximum sensi- 
tivity to red light not where most of 
the phytochrome is found, namely in 
the node itself and in the coleoptile tip, 
but rather just below the node (for the 
mesocotyl response) or both just below 
and just above the node (for the coleop- 
tile response). She also discovered that 
there is significant light piping through 
both organs, with fairly sharp attenu- 
ation across the node. By quantifying 
this light piping, she was able to account 
in part for the dose-response curve on the 
basis of reactions of the two photoper- 
ceptive sites to light directly absorbed at 
the sites as well as that piped there from 
other regions of the seedling. These find- 
ings are of considerable importance, par- 
ticularly in view of the magnitude of the 
piping and the fact that the possibility of 
light piping has been virtually ignored in 
all previous studies. 

Walton and Shinkle have made signi- 
ficant progress in understanding the 
mechanism by which the corn meso- 
cotyl responds to low fluences of red 
light. Walton had already shown that 
red light strongly reduced the amount 
of the glucan synthetase localized in 
the Golgi in corn mesocotyls. In the 
present work, he and Shinkle have 
shown that the fluence-response curve 
for reduction of the glucan synthetase 
activity was almost identical to that 
for mesocotyl growth suppression. 
Both effects were somewhat reversible 
by far-red light, and both could be par- 
tially potentiated by far-red light 
alone. Auxin, which Vanderhoef and 
Briggs showed some years ago would 
antidote the growth suppression caused 
by red light, also antidotes the decline 
in glucan synthetase activity. Hence 
both growth and glucan synthetase ac- 

tivity may well be modulated by a red- 
light-regulated supply of auxin. 

Leong, continuing his detailed char- 
acterization of the light-sensitive fla- 
vin-cytochrome b complex in corn mem- 
brane preparations, has shown its 
association with a particular ATPase, 
quantified the amount of flavin, and 
measured the midpoint potential of the 
cytochrome. Since the system is al- 
most certainly located in the plasma 
membrane, knowledge of these prop- 
erties should provide considerable in- 
sight into the functional nature of the 
membrane system limiting the plant 

Until this year, there was little evi- 
dence linking this pigment complex 
with photoreception for phototropism 
in corn, save that both light-induced 
cytochrome reduction and phototrop- 
ism were inducible by blue light. Re- 
cently, however, Leong has made con- 
siderable progress by using a new class 
of herbicides based on a diphenyl ether 
structure. He showed that these com- 
pounds (which require light for their 
herbicidal action) significantly enhance 
photoreduction of the cytochrome. 
They also significantly sensitize the 
phototropic response of etiolated oat 
seedlings! These experiments substan- 
tially strengthen the case that the fla- 
vin-cytochrome complex does indeed 
serve as the photoreceptor for photo- 
tropism. Further, the herbicides pro- 
vide a specific probe which may help us 
to elucidate the function of this light- 
sensitive electron transport chain in 
the plant plasma membrane. 

The final report brings us full circle, 
back to the computer and the RE SOL 
program. Schoch, Gorton, and Briggs 
reinvestigated the dramatic spectral 
changes, first described by Shibata in 
this Department some 25 years ago, 
which occur after protochlorophyll 
phototransformation in the leaves of 
dark-grown seedlings. These changes 
involve a shift in the absorption maxi- 
mum of the newly formed chlorophyll 
from about 684 nm to about 670 nm. 



Schoch and her co-workers were able to 
show (1) that carotenoids, despite their 
extensive presence in the proplastids, 
were in no way involved in the shift, (2) 
that reduction of the double bonds of 
the long aliphatic tail of chlorophyll 
was likewise not involved, and (3) that 
the shift probably did not involve pro- 
teolytic cleavage of the apoprotein. 
While we are still unaware of the physi- 

cal or chemical changes behind the 
shift, we can eliminate two possibilities 
and consider a third unlikely. 

The findings briefly described in 
these few paragraphs are far better doc- 
umented in the reports that follow, and 
virtually all of them either have ap- 
peared or will soon appear in the pub- 
lished literature. They represent a fine 
record of high-caliber productivity. 

Winslow R. Briggs 


Jeanette S. Brown and Siegrid Schoch 

Investigation of the state of chloro- 
phyll in photosynthetic membranes is 
of major interest to the researchers of 
this Department. Some years ago, 
French (French et al., 1972) proposed 
that chlorophyll a exists in four major 
and several minor different states in 
vivo. This hypothesis was based on the 
results of curve analyses of absorption 
spectra of many plant species. For the 
most part, the spectra were from par- 
ticles prepared by forcing chloroplasts 
or algae through the needle-valve of a 
French pressure cell, a treatment 
which does not usually alter the nature 
of the pigment. Procedures have re- 
cently been developed to isolate the 
more elementary chlorophyll-protein 
complexes yet retaining the pigments 
in their native state. 

Our analyses of complex spectra 
have been greatly facilitated this year 
by two major improvements. A shaft 
encoder and microprocessor connected 
to our Cary 17 spectrophotometer 
make possible the transfer of absorp- 
tion spectra to a Hewlett-Packard com- 
puter system. In addition, Mr. Glenn 
Ford has slightly modified the RE SOL 
{Year Book 67, 536-546) program to 
run on this computer. It takes only 
15-20 min to perform a complete anal- 
ysis of a spectrum having 2048 digital 
points and a maximum of nine compo- 

nents, allowing ten iterations. Because 
each curve deconvolution is relatively 
brief and inexpensive, we are able to 
analyze each spectrum more carefully 
and to compare many more spectra 
than previously {Year Book 70, 

With these improvements in analyti- 
cal procedures, we have begun a pro- 
gram to analyze spectra of native chlo- 
rophyll-protein complexes from many 
sources. We expect to test French's 
original hypothesis further and to ex- 
tend our knowledge of the way chloro- 
phyll functions in photosynthesis. 

We isolated a light-harvesting chlo- 
rophyll a-6-protein (LHCP) complex 
and photosystem I (PS I) complex 
from both wheat and pea chloroplasts 
using Triton X-100 according to the 
procedure of Burke et al. (1978). De- 
tails of these experiments and results 
of the curve analyses of the absorption 
spectra are given in Brown and Schoch 

All of the spectra could best be fitted 
with the same four major component 
bands found by French et al. (1972). As 
also observed by these same authors, 
the half-band widths of the compo- 
nents in the PS I complexes were con- 
sistently greater (by 1-3 nm) than in 
the LHCP. The amounts of the 660-nm 
and 678-nm components were propor- 



tionately the same in all the spectra, 
suggesting that exciton interaction 
may cause an absorption band split- 
ting between these two components. 

The addition of a narrow band at 
675-676 nm comprising about 1% of 
the total absorption remarkably im- 
proved the overall curve fitting of the 
LHCP and chloroplast spectra without 
changing the other bands, but it did 
not improve the analysis of the PS I 
spectra. The significance or possible 
function of this small component is un- 

It is known from other kinds of 
studies (Thornber et al, 1977) that 
LHCP comprises about 50-60% of the 
total chlorophyll in higher plants, and 
the PS I fraction about 10-25%. Al- 
though individual analyses of the 
wheat or pea LHCP and PS I spectra, 
compared to analyses of spectra of 
chloroplast particles, indicated no de- 
struction of chlorophyll forms by the 
fractionation procedure, we could not 
fit the chloroplast spectra with the 
sum of spectra of these two fractions in 
any reasonable proportion. This result 

indicated that at least one major chlo- 
rophyll fraction was not recovered by 
our procedure. We calculated the shape 
of this missing fraction, which has a 
major component band absorbing near 
684 nm. Other circumstantial evidence 
(Delepelaire and Chua, 1979; Waldron 
and Anderson, 1979) suggests that this 
missing fraction may contain the chlo- 
rophyll a antenna of photosystem II. 
These results indicate that spectral 
analysis of isolated chlorophyll-protein 
complexes will continue to aid in 
understanding the state of chlorophyll 
in photosynthetic membranes. 


Brown, J. S., and S. Schoch, Biochim. Biophys. 

Acta, 636, 201-209, 1981. 
Burke, J. J., C. L. Ditto, and C. J. Arntzen, 

Arch. Biochem. Biophys., 187, 252-263, 1978. 
Delepelaire, P., and N. H. Chua, Proc. Nat. Acad. 

Sci. USA, 76, 111-115, 1979. 
French, C. S., J. S. Brown, and M. C. Lawrence, 

Plant Physiol, 49, 421-429, 1972. 
Thornber, J. P., R. S. Alberte, F. A. Hunter, 

J. A. Shiozawa, and K.-S. Kan, Brookhaven 

Sym. Biol., 28, 132-148, 1977. 
Waldron, J. C, and J. M. Anderson, Eur. J. 

Biochem., 102, 357-362, 1979. 




Siegrid Schoch and Jeanette S. Brown 

Striking differences have been ob- 
served in the fluorescence emission and 
excitation spectra of chlorophyll a be- 
tween normally greened plants and 
plants grown under light-limiting con- 
ditions (Year Book 79, 176-179; Davis 
et al, 1976; Strasser and Butler, 
1977a,b; Mullet et al, 1980). Plants 
developed under light-limiting condi- 
tions form very little chlorophyll b and 
essentially none of the light-harvesting 
chlorophyll a-b protein (LHCP) (Argy- 
roudi-Akoyunoglou and Akoyunoglou, 
1970; Ryberg et al, 1980). Both photo- 
systems are fully active in these plants, 
and they offer the possibility of mea- 

suring the spectroscopic properties of 
chlorophyll a without interference from 
the normally large proportion of LHCP. 
Last year, we compared the fluores- 
cence of fully green, daylight-grown 
(DL) and intermittent-light-grown 
(ImL), young wheat and pea plants. 
Now, with the new computer connec- 
tion to the Cary spectrophotometer (see 
Brown and Schoch, this Report), we 
have extended these studies to absor- 
bance measurements. 

Wheat seedlings were cultivated in 
the same manner as last year except 
that the ImL-grown plants were illu- 
minated for 15 min out of 3 hours 



darkness instead of 2 hours. This 
change decreased the possibility of 
LHCP formation even more. Chloro- 
plasts were also prepared in the same 
manner as last year from both DL- 
grown and ImL-grown, seven-day-old 

Absorption spectra were measured 
at 89 K in a Cary spectrophotometer 
equipped with a scattering transmis- 
sion attachment and Dewar cell holder 
(Brown and Schoch, 1981). Computer 
analysis of these spectra was also car- 
ried out, as described elsewhere (Brown 
and Schoch, 1981). 

Resolved spectra of DL-grown and 
ImL-grown wheat chloroplasts are 
shown in Fig. 1. The data from these 
two analyses are given in the first two 
columns of Table 1. The sum of compo- 
nents listed in Table 1 for the DL 
chloroplasts gave an excellent, rela- 
tively unique fit to the experimental 
spectrum, but several sets of some- 
what different components could be 
used to fit the ImL spectrum with 
nearly the same precision. The last col- 
umn of Table 1 lists the components 
resolved from a spectrum of wheat 
LHCP prepared from green chloro- 
plasts (Brown and Schoch, 1981). 

Figure 2 shows a difference spectrum 
of DL-minus-ImL spectra. These spec- 
tra were normalized before subtraction 
to make the greatest difference near 
650 nm, the chlorophyll b region, be- 
cause we know that the amount of chlo- 
rophyll b is indeed much lower in the 
ImL chloroplasts. With this qualifica- 
tion, the spectra had equal absorbance 
at 668 nm. The green chloroplasts had 
more absorption near 660, 678, and 690 
nm, as well as in the chlorophyll b re- 
gion, whereas the ImL plastids showed 
a unique peak at 683 nm. 

Although the choice is admittedly 
somewhat subjective, we prefer to com- 
pare the third analysis (Table 1, col- 
umn 4) of the ImL spectra with that of 
the DL chloroplasts because a rela- 
tively close (error X 16) fit was ob- 
tained with only seven components. 

The significant differences were as 

1. The component thought to repre- 
sent chlorophyll b near 650 nm is not 
as reduced in the ImL spectra (from 21 
to 17) as might be expected. If the 
639-nm band is included with the 650- 
nm band, the difference becomes 
greater (from 41 to 26 or from 55 to 26) 
when compared with the chlorophyll b- 
enriched LHCP. The results suggest 
that up to half of the 650-nm compo- 
nent in LHCP-containing chloroplasts 
may include absorption by lower vibra- 
tional levels of the longer-wavelength 
chlorophyll a components as well as by 
chlorophyll b. 

2. The proportions of the 661-nm 
and 678-nm components were the same 
in DL and ImL spectra, while the 
670-nm component increased in the 
ImL spectrum. This observation sup- 
ports the suggestion of exciton interac- 
tion between the 661-nm and 678-nm 
forms of chlorophyll (Brown and 
Schoch, 1981). Also, as predicted, the 
addition of a small 676-nm band did 
not improve the fit in ImL spectra. 

3. A 683-nm component in the ImL 
spectrum is prominent in the difference 
spectrum (Fig. 2) and is different from 
the 683 band in normal green spectra. 
As shown in Table 1, this ImL 683 
band has a much narrower width (by 5 
nm) and is over 50% more Lorentzian 
in shape. A similar 683-nm component 
was previously observed in spectra of 
an isolated chlorophyll a-protein com- 
plex from Euglena {Year Book 73, 
694-706). Fluorescence excitation 
spectra of ImL chloroplasts but not of 
DL chloroplasts {Year Book 79, 
176-179) show a peak near 683 nm. Cir- 
cumstantial evidence links this 683 
component with photosystem II (Dele- 
pelaire and Chua, 1979). Perhaps this 
special band is masked by antenna 
chlorophyll in fully greened plastids. 

4. There is little or no absorption 
beyond 695-nm by ImL-grown plants. 
Also, the 689-nm component in the 
ImL spectrum was 87% Lorentzian 








>-! h 




«3 H 

or X ] 


qw a 

xA W 



(M ,-h 



. o 



w £ 

en W 

Tf Tf H rt 1> O N 

qo co n n ; o c- cn 


tJ ^r CM CO CO O) C7) 00 

■^■Tf-COCN-rfCNCOCN . t- 

i-H (M r-l Ol -Tf" CO 

lO COt- [- 00 <7> 


co i i i t- i i i o o 


-tf CO C- t- CO 00 

co co CO CO CO CO 

h a 

fn CCS 


?> « 





compared to 100% Gaussian in the DL the decrease in long-wavelength fluo- 

spectrum. This difference in and/or rescence emission observed earlier 

lack of long-wavelength-absorbing (Year Book 79, 176-179; Davis et al, 

forms of chlorophyll can account for 1976; Mullet et al, 1980). 






B. ImL 

'avelencjth, n 


Fig. 1. Absorption spectra at 89 K of wheat chloroplasts grown under daylight (A) or inter- 
mittent light (B). The measured data are plotted as points, while the line through them is the sum of 
the component curves, the characteristics of which are given in Table 1. The error of fit at each point 
is shown below each spectrum on a scale with the designated magnification. 



650 670 690 710 730 

Wavelength, nm 

Fig. 2. Differences between absorption spec- 
tra of green and ImL chloroplasts from wheat. 
The spectra were normalized at 668 nm before 

These four major differences, found 
by comparing detailed analyses of chlo- 
rophyll spectra of the same plant spe- 
cies grown under different light condi- 

tions, show the value of such studies 
for our eventual understanding of the 
native state of chlorophyll and how it 
functions in photosynthesis. 


Argyroudi-Akoyunoglou, J. H., and G. Ako- 

yunoglou, Plant Physiol, 46, 247-249, 1970. 
Brown, J. S., and S. Schoch, Biochim. Biophys. 

Acta, 636, 201-209, 1981. 
Davis, D. J., P. A. Armond, E. L. Gross, and 

C. J. Arntzen, Arch. Biochem. Biophys., 175, 

64-70, 1976. 
Delepelaire, P., and N. H. Chua, Proc. Nat. Acad. 

Sci. USA, 76, 111-115, 1979. 
Mullet, J. E., J. J. Burke, and C. J. Arntzen, 

Plant Physiol., 65, 823-827, 1980. 
Ryberg, H., L. Axelsson, K. -O. Widell, and H. I. 

Virgin, Physiol., Plant, 49, 431-436, 1980. 
Strasser, R. J., and W. L. Butler, Biochim. 

Biophys. Acta, 462, 295-306, 1977a. 
Strasser, R. J., and W. L. Butler, Biochim. 

Biophys. Acta, 462, 307-316, 1977b. 




Kazuhiko Satoh 

It is well known that when dark- 
adapted algal cells or leaves are sud- 
denly illuminated with strong light, 
the yield of chlorophyll fluorescence 
shows several transients before it 
reaches a steady-state level (Kautsky 
and Appel, 1960; Govindjee and Papa- 
georgiou, 1971; Katoh et al., 1975). The 
most pronounced transient in the in- 
duction of chlorophyll fluorescence is 
the DPS^ which occurs concomitantly 
with the induction of cytochrome /pho- 
tooxidation. Satoh et al (1977) ex- 
plained this DPSi transient and the in- 
duction of cytochrome / oxidation as 
reflecting a photoactivation on the re- 
ducing side of photosystem I. 

Recently, Satoh and Katoh (1980), 
using intact spinach chloroplasts, ob- 
served that nitrite also decreased the 
DPSi transient. Nitrite accepts elec- 
trons from reduced ferredoxin through 

nitrite reductase (Heber and Purczeld, 
1977). The results indicate that the 
dark inactivated site is on the reducing 
side of ferredoxin, because nitrite re- 
duction bypassed the inactivated site. 
Satoh and Katoh (1980) also showed 
that oxalacetate and 3-phosphoglycer- 
ate, both of which accept electrons 
from NADPH, had little effect on the 
DPS! transient. From measurements 
of the activity of glyceraldehyde 3- 
phosphate dehydrogenase and malic 
dehydrogenase, they concluded that 
the regulation site was located before 
the reduction of NADP+. Ferredoxin- 
NADP + reductase is the only enzyme 
occurring between ferredoxin and 
NADP+ and, therefore, ferredoxin- 
NADP+ reductase might be the en- 
zyme subject to changes of activity 
during a dark-light transition. 
In this work, we used isolated intact 



chloroplasts from Bryopsis and estab- 
lished conditions under which the enve- 
lopes of the chloroplasts became leaky 
without affecting the DPSi transient. 
Under such conditions, we were able to 
observe the change in the activity of 
the enzyme ferredoxin-NADP + reduc- 
tase during the dark-to-light transition 
of the chloroplasts. We also found that 
reduction of the enzyme might induce 
the photoactivation and that oxygen 
acted as a principal electron acceptor 
after the photoactivation had taken 


Studying the cause of the DPSi tran- 
sient was difficult because the tran- 
sient could be observed only in cells or 
in intact chloroplasts which had bar- 
riers to almost all ions and even small 
molecules. Therefore, we tried to get 
leaky chloroplasts by changing the os- 
molarity of the reaction medium. Fig- 
ure 3 shows the rates of C0 2 fixation 
and NADP+Hill reaction at various 
concentrations of sorbitol in the reac- 
tion mixture. In the case of the 
NADP+-HU1 reaction, 1 raM NADP+ 
was added to the reaction mixture and 
absorbance changes at 340 nm were re- 
corded. At 1.0 M sorbitol, Bryopsis 
chloroplasts showed a high rate of C0 2 
fixation but no NADP+-HH1 reaction. 
This is because NADP+ cannot pene- 
trate into the chloroplasts in intact 
Bryopsis chloroplasts. The photo- 
reduction of intrinsic NADP + was 
probably too small to be measured by 
our technique. Upon lowering the con- 
centration of sorbitol, the rate of the 
CO 2 fixation reaction was decreased. 
Concomitantly, the rate of the NADP + - 
Hill reaction increased. At 0.25 M sor- 
bitol, the CO 2 fixation was greatly 
depressed and the rate of NADP+Hill 
reaction reached a relatively high level. 
The increase of NADP+-Hill reaction and 
the decrease of the C0 2 fixation show that 
the envelopes of the chloroplasts became 
leaky only to small molecules such as 

NADP+, and that large molecules such as 
ferredoxin may have remained in the chlo- 
roplasts, thus explaining the high rate of 
the NADP + -Hill reaction at low concen- 
trations of sorbitol. 

Figure 4 shows the time course of 
chlorophyll fluorescence at various 
concentrations of sorbitol. The marked 
DPSj transient remained unaffected at 
a concentration of 0.25 M sorbitol (Fig. 
4b) but it became less pronounced at 
concentrations of 0.1 M or less (Fig. 
4c,d). The persistence of the DPSj tran- 
sient (despite the loss of the C0 2 fixa- 
tion at low concentrations of sorbitol) 
supports the suggestion that the C0 2 
fixation reaction is not required for the 
DPS! transient. 

Figure 4 shows that addition of 1 
mM NADP+ had no significant effect 
on the DPS! transient at 0.25 M sor- 
bitol, suggesting that the availability 
of NADP + is not a requirement for the 
DPSi. Only when the sorbitol concen- 
tration was lowered below 0.1 M did the 
effect of NADP+ become noticeable 
(Fig. 4f,g). In the absence of sorbitol, 
NADP + markedly decreased the ex- 
tent of chlorophyll fluorescence, indi- 
cating that the functional integrity of 
the system may have been damaged. 

Ferredoxin-NADP + reductase is 
known to catalyze the reduction of 
DCIP by NADPH in the dark (Shin, 
1971). Therefore, if this enzyme 
changes its activity in the dark and 
light conditions, the diaphorase activ- 
ity may also change according to the 
dark-to-light transition of the chloro- 
plasts. The diaphorase activities were 
measured at two sorbitol concentra- 
tions, 0.25 M and M. At 0.25 M, a 
marked difference in the rate was ob- 
served in dark- and light-adapted 
chloroplasts. But at M sorbitol, little 
difference was observed (Table 2). 

To elucidate the mechanism of the 
photoactivation, we measured the ef- 
fects of incubation of the chloroplasts 
in the dark with reductants at 0.25 M 
sorbitol concentration (Fig. 5). The 
reductants were removed by two cen- 




















Concentration of sorbitol, M 

Fig. 3. Effects of concentration of sorbitol in the reaction medium on the rate of photosynthesis 
and NADP + -Hill reaction in Bryopsis chloroplasts. The reaction mixture contained in 2 ml, 50 mM 
HEPES (pH 7.5), 11 mM MgCl 2 , 1 mM MnCl 2 , 2 mM NaN0 3 , 2 mM EDTA, and various concentra- 
tions of sorbitol, as indicated in the figure. Chlorophyll concentrations were 26.2 /xg/ml in the 
photosynthesis measurements (circles) and 10.5 Mg/ m l i n the NADP + -Hill reaction measurements 

trifugations of the chloroplasts. Since 
the dark inactivation process is very 
slow in Bryopsis chloroplasts (data not 
shown), we can see whether the activa- 
tion of electron transport has taken 
place if we measure the induction of 
chlorophyll fluorescence less than 10 
min after the removal of the reduc- 
tants. The time course of chlorophyll 
fluorescence in the chloroplasts that 
had been incubated with ascorbate 
(Fig. 5b) was similar to that in control 
chloroplasts which had been incubated 
with no reductant (Fig. 5a). Incubation 

of the chloroplasts with dithiothreitol 
slightly decreased the DPSi transient 
(Fig. 5c). A dramatic effect was ob- 
tained by the addition of dithionite 
(Fig. 5d). Incubation of the chloro- 
plasts with dithionite largely elimi- 
nated the DPSx transient of chloro- 
phyll fluorescence, indicating that the 
photoactivation process had already 
taken place. These results show that 
the reduction of a certain substance, 
which can be reduced by dithionite and 
partly by dithiothreitol but not by 
ascorbate (i.e., ferredoxin-NADP + re- 



TABLE 2. Diaphorase Activities in Dark- and 
Light-Adapted Chloroplasts. 

Rates of DC IP Reduction 

(relative units) 




0.25 M 3.0 (±0.5)* 

0M 7.4 (±1.1) 

5.5 (±0.9) 
7.8 (±1.0) 

*Numbers in parentheses are standard devia- 
tions. Each value is the average of five measure- 
ments. Chlorophyll concentration was 14.8 
/ig/ml. Other conditions were the same as in 
Fig. 3. 

0. 25M 

0. 1M 

Fig. 4. Effects of concentration of sorbitol 
and addition of NADP + on the DPSj transient 
of chlorophyll fluorescence: (a) 1.0 M sorbitol; (b) 
and (e) 0.25 M sorbitol; (c) and (f) 0.1 M sorbitol; 
(d) and (g) M sorbitol. 1.0 mM NADP+ was 
also added in (e), (f), and (g). The intensity of the 
excitation light was 2.0 X 10 4 erg cm -2 s _1 . 
Chlorophyll concentration was 5.41 uglml. Other 
conditions were the same as in Fig. 3. 






Fig. 5. Time courses of chlorophyll fluores- 
cence after incubation of the chloroplasts with 
ascorbate, dithiothreitol, or dithionite at 0.25 M 
sorbitol. Chloroplasts were incubated for 5 min 
with no reductant (a), with 10 mM ascorbate (b), 
with 10 mM dithiothreitol (c), or with 5 mM 
dithionite (d). The incubation medium contained 
0.25 M sorbitol, 50 mM HEPES (pH 7.5), 2 mM 
EDTA, 11 mM MgCl 2 , 1 mM MnCl 2 , 2 mM 
NaN0 3 . The reductants were removed by 
washing the chloroplasts with two centrifuga- 
tions (see text). Chlorophyll concentrations were 
2.86, 2.44, 2.59, and 2.31 /xg/ml for curves a, b, c, 
and d, respectively. Other conditions were the 
same as in Fig. 4. 

ductase), is involved in the photoacti- 
vation process. 

The photoactivation process is also 
known to be related to oxygen. In the 
absence of oxygen, the DPSj transient 
was inhibited (Bannister and Rice, 
1968). Oxygen may participate in the 
photoactivation process in two ways. 
First, it may act as a mediator; in other 
words, oxygen may be necessary for 
the process itself. Second, it may act as 
an electron acceptor after the photoac- 
tivation process has taken place. In 
this case oxygen is not necessary for 
the process itself. Therefore, we tested 
whether the photoactivation process 
would occur in the absence of oxygen. 
Chloroplasts were preilluminated for 
10 sec in the absence of oxygen (in the 
presence of 5 mM dithionite) and then 
washed twice by centrifugation to re- 
move the dithionite. The fluorescence 




L J 



Fig. 6. Effects of anaerobiosis on the fluorescence yield after preillumination of the chloroplasts. 
10.0 mM /3-D-glucose and 60 units/ml of catalase were added. In curve a, chloroplasts were dark 
adapted. In curves b and c, chloroplasts were preilluminated with the actinic light for 10 s, 3 min 
before the measurements. In curve c, 20 units/ml of glucose oxidase was added 30 s after the 
preillumination of the chloroplasts. Chlorophyll concentration was 4.62 /xg/ml. Other conditions were 
the same as in Fig. 3. 

time course of the chloroplasts showed 
no DPSx transient, indicating that the 
photoactivation process had already 
taken place (data not shown). Addition 
of dithionite without preillumination 
had little effect on the DPSi transient 
in 1.0 M sorbitol. These data show 
clearly that oxygen is not necessary for 
the photoactivation process itself, thus 
eliminating the first possibility. 

In order to test whether oxygen acts 
as an electron acceptor after chloro- 
plasts are photoactivated, the yield of 
chlorophyll fluorescence was measured 
under anaerobic conditions after photo- 
activation. In the presence of oxygen 
and after the photoactivation, elec- 
trons flow rapidly, Q remains oxidized 
and, therefore, the fluorescence yield 
remains low (Fig. 6b). In the absence of 
oxygen, however, the extent of chloro- 
phyll fluorescence reached higher lev- 
els, showing that Q was largely re- 
duced in photoactivated chloroplasts 
(Fig. 6c). The extent and time course of 
this fluorescence change was the same 
as those for nonphotoactivated chloro- 
plasts under anaerobic conditions, in- 
dicating that the difference between 
the yields of chlorophyll fluorescence 
in Fig. 6b and 6c is not due to the 

quenching of chlorophyll fluorescence 
by oxygen (data not shown). These 
data show that even after the photo- 
activation of ferredoxin-NADP + re- 
ductase, electrons do not flow rapidly 
under anaerobic conditions, and that 
oxygen is acting as an electron accep- 


Bannister, T. T., and G. Rice, Biochim. Biophys. 
Acta, 162, 555-580, 1968. 

Govindjee, and G. Papageorgiou, in Photophysi- 
ology, Vol. 6, 1-46, A. C. Giese, ed., Academic 
Press, New York, 1971. 

Heber, U., and P. Purczeld, in Proc. Fourth Intl. 
Congr. ofPhotosyn. Res., 107-118, D. O. Hall, 
J. Coombs, and T. W. Goodwin, eds., The Bio- 
chemical Society, London, 1977. 

Katoh, S., K. Satoh, A. Yamagishi, and T. Ya- 
maoka, Plant Cell Physiol, 16, 1093-1099, 

Kautsky, H., and W. Appel, Biochem. Z., 332, 
277-292. 1960. 

Satoh, K., A. Yamagishi, and S. Katoh, in Photo- 
synthetic Organelles, Special Issue of Plant 
Cell PhysioL, 75-86, S. Miyachi, S. Katoh, 
Y. Fujita, and K. Shibata, eds., Japanese Soci- 
ety of Plant Physiologists and Center of Aca- 
demic Publications Japan, Tokyo, 1977. 

Satoh, K., and S. Katoh, Plant Cell Physiol., 21, 
907-916, 1980. 

Shin, M., in Methods in Enzymology, Vol. 23, 
440-447, A. San Pietro, ed., Academic Press, 
New York, 1971. 






Anastasios Melis and Geoffrey W. Harvey 

The use of a sensitive spectrophoto- 
metric method has recently allowed 
workers to determine the stoichiome- 
tric ratio of PS II to PS I reaction cen- 
ters in different photosynthetic mem- 
branes (Melis and Brown, 1980). It was 
found that considerable differences in 
this ratio existed between membranes 
from the grana and stroma regions of 
chloroplasts. Grana membranes con- 
tained most of the chlorophyll 6, photo- 
system II centers, and plastoquinone 
of the chloroplast. Stroma membranes 
contained primarily photosystem I 
reaction centers. In the present in- 
vestigation, we used different spectral 
qualities of light during plant growth 
to probe the relationship between cer- 
tain chloroplast structural and func- 
tional parameters. We found that the 
quality of the absorbed light controlled 
the relative concentrations of system 
II and system I reaction center com- 
plexes and, in addition, it controlled 
the chloroplast membrane differentia- 
tion into grana and stroma lamellae. 


A first group of experiments was 
conducted by growing Pisum plants in 
the laboratory under continuous illumi- 
nation of different spectral qualities. 
Table 3 compares the results obtained 

with Pisum chloroplasts developed in 
the greenhouse (control), under far-red- 
deficient (cool white fluorescent) light, 
or far-red-enriched (incandescent) light. 
Greenhouse and far-red-deficient plants 
showed minor quantitative differences 
in the Chi a/Chl b ratio, and in their Q 
and P 70 o contents. However, chloro- 
plasts developed in far-red-enriched 
light showed a significant quantitative 
difference from those developed in far- 
red-deficient light in the Q and P 70 o 
contents: on a chlorophyll basis we con- 
sistently measured an enrichment in the 
far-red-developed chloroplasts in the 
amount of Q present and a simultane- 
ous decrease in the concentration of 
P700 Such alterations shifted the 
Q/P700 ratio from the average value of 
1.8 to 2.8, representing an overall 
change of approximately 60%. 

Figure 7 compares, in cross section, 
the ultrastructure of Pisum chloro- 
plasts developed under far-red-defi- 
cient and far-red-enriched light. There 
is a considerable difference in the rela- 
tive abundance and size of grana and 
stroma lamellae in the two samples. 
Under far-red-deficient light the chloro- 
plasts appeared to have thinner grana 
stacks and more extended stroma thy- 
lakoids (Fig. 7, upper). Far-red-en- 
riched light generally yielded a higher 
density of thicker grana (Fig. 7, lower). 


Chlorophyll a to Chlorophyll 

b Ratios, Q Content 

and P700 Content in Pisum 

Chi a/Chl b 

Q, /xmol 

P 700- M mo1 



2.50 ± 0.05 
2.96 ± 0.05 
2.64 ± 0.05 

0.42 ± 0.06 
0.40 ± 0.06 
0.49 ± 0.06 

0.23 ± 0.01 
0.22 ± 0.01 
0.17 ± 0.01 


*The amounts of Q and P 700 in ^mol correspond to 100 ^mol chlorophyll (a + b). The chloroplasts 
were developed either in the greenhouse, by far-red-deficient cool white fluorescent light, or by far- 
red-enriched incandescent light. 



Fig. 7. Electron micrographs of Pisum chloroplasts greened under far-red-deficient (upper) and 
under far-red-enriched light (lower). 



Fig. 8. Electron micrographs of Asarum (upper) and Polystichum (lower) chloroplasts showing 
the extended grana stacks, high membrane density, and decreased stroma volume, typical of shade- 
adapted plants. 



Since grana thylakoids are enriched in 
Q (Melis and Brown, 1980), the above 
structural changes are in agreement 
with the light quality-induced changes 
in Q and P 70 o contents. 

A second group of experiments in- 
volved shade species, which in nature 
occur only in the lower vegetation level 
of densely shaded habitats. An impor- 
tant feature of shade plants is their 
chloroplast size and ultrastructure 
(Bjorkman, 1973; Boardman et al, 
1974). Figure 8 shows the chloroplast 
ultrastructure of the typical shade spe- 
cies Asarum and the fern Polystichum, 
revealing the large well-developed 
grana stacks, which in many cases ex- 
tend across the entire chloroplast 
body. A concomitant decrease in the rel- 
ative number or length of the inter- 
grana stroma lamellae was not always 
apparent in the shade species we ex- 
amined. However, the density of the 
membrane phase (grana and stroma 
thylakoids) was always higher than in 
chloroplasts from plants exposed to 
full sunlight, occupying almost the en- 
tire chloroplast volume and thus re- 
sulting in a drastic reduction of the rel- 
ative stroma volume (Boardman et al, 
1974). This result correlates with the 
decreased levels of ribulose bisphos- 
phate carboxylase and other soluble 
proteins (Bjorkman, 1968) in such chlo- 
roplasts. Interestingly, in the three 
shade species examined, the light-satu- 
rated uncoupled rates of electron trans- 
port were low, ranging between 100 
and 250 /xeq mg Chi -1 h -1 (Boardman 
et al, 1974; Boardman, 1977). Table 4 
also shows that in the shade species 
Asarum, Polystichum, and Tolmiea the 

Chi a/Chl b ratios were considerably 
lower and the stoichiometric ratios of 
Q/P 7 oo were generally higher than that 
of sun-adapted plants. (Compare with 
data in Table 3.) In these chloroplasts 
there is a parallel increase in Chi b, PS 
II content, and thylakoid membrane 
density in the chloroplast volume. 

Since a shade habitat is an environ- 
ment of low light intensity but is 
enriched in far-red light (Bjorkman, 
1973), it may be observed that chloro- 
plasts from the above shade species 
show the combined effect of the low 
light intensity on the Chi a/Chl b ratio 
(Boardman, 1977) superimposed on the 
light quality effect on the reaction cen- 
ter ratio (Q/P700K 

Under predominantly far-red-enriched 
illumination, PS I reaction centers turn 
over electrons faster than their PS II 
counterparts. The increased PS I I/PS I 
ratio observed under these conditions 
may serve to offset this energetic 
imbalance and to maintain a balanced 
electron flow through the electron 
transport chain. 

The quality of light during chloro- 
plast growth apparently controls both 
the differentiation of the membrane 
phase into areas of grana stacks and 
stroma thylakoids and the stoichiome- 
tric amounts of PS II and PS I reac- 
tion centers. It may be concluded that 
in oxygen-evolving organisms, PS II 
and PS I constitute integral thylakoid 
membrane-protein complexes that are 
structurally and, therefore, stoichiome- 
trically independent of each other. We 
hypothesize that the structural dif- 
ferentiation of the chloroplast mem- 
branes into grana regions occurs, at 

TABLE 4. Chlorophyll a to Chlorophyll 
Adapted Species Asarum candatum, 

6 Ratios, Q Content, and P700 Content in the Shade- 
Polystichum munitum, and Tolmiea menziesii* 

Chi a/Chl b 

Q, ^mol 

P 700 , (umol 





2.32 ± 0.4 
2.29 ± 0.05 
2.31 ± 0.05 

0.52 ± 0.07 
0.62 ± 0.07 
0.46 ± 0.07 

0.19 ± 0.01 
0.16 ± 0.01 
0.19 ± 0.01 


*The amounts of Q and P 700 in moles correspond to 100 mol chlorophyll (a + b). 



least in part, in order to facilitate the 
centralization and relative concentra- 
tion increase of PS II reaction centers. 
This subject is discussed more fully 
elsewhere (Melis and Harvey, 1981). 


Bjorkman, O., Physiol. Plant, 21, 1-10, 1968. 
Bjorkman, O., in Photophysiology, Vol. 8, 1-63, 

A. C. Giese, ed., Academic Press, New York, 

Boardman, N. K., J. M. Anderson, O. Bjorkman, 

D. J. Goodchild, L. H. Grimme, and S. W. 

Thome, Port. Acta Biol, 14, 213-236, 1974. 
Boardman, N. K., Annu. Rev. Plant Physiol, 28, 

355-372, 1977. 
Melis, A., and J. S. Brown, Proc. Nat. Acad. Sci. 

USA, 77, 4712-4716, 1980. 
Melis, A., and G. W. Harvey, Biochim. Biophys. 

Acta, in press, 1981. 






Eduardo Zeiger, * Paul Armond, and Anastasios Melis 

Chloroplasts are a central structural 
feature of stomatal guard cells (Meid- 
ner and Mansfield, 1968). Many invest- 
igators have suggested ways in which 
these chloroplasts might be crucial for 
stomatal functioning (Zelitch, 1969; 
Hsiao, 1976), but most hypotheses 
have proven untenable in the face of 
new discoveries or have not received 
conclusive experimental support. A bet- 
ter understanding of the physiological 
and biochemical properties of guard 
cell chloroplasts is crucial for a defini- 
tion of their role in stomatal move- 
ments. Further progress depends largely 
on our ability to obtain pure prepara- 
tions of guard cells and to monitor the 
photosynthetic activity of their chloro- 
plasts in vivo. 

Here, we used albino portions of var- 
iegated leaves from Chlorophytum com- 
osum as a source of uncontaminated 
guard cell chloroplasts and studied 
their pigment content by fluorescence 
spectroscopy. Our findings indicate 
that guard cell chloroplasts contain 
light-harvesting pigments of both PS I 
and PS II. The fluorescence induction 
kinetics of guard cell and mesophyll chlo- 
roplasts are also presented. Our results 

* Department of Biological Sciences, Stanford 
University, Stanford, CA 94305. 

support the notion that guard cell chlo- 
roplasts operate a linear electron trans- 
port system. We postulate that a cen- 
tral role of guard cell chloroplasts is to 
provide light-dependent energy to sus- 
tain the active ion transport required 
for stomatal opening during the day. 


Emission Spectra of Guard Cell Chloro- 
plasts. Low-temperature (77 K) emission 
spectra were obtained in a fluoro- 
spectrophotometer from samples frozen 
in liquid N 2 . Because of the high sensi- 
tivity of this technique, it was crucial to 
use preparations of guard cell chloro- 
plasts free from contaminating meso- 
phyll. The albino portions of variegated 
leaves from C comosum provided an 
adequate source. An albino leaf seg- 
ment, previously examined under fluo- 
rescence microscopy to ensure the ab- 
sence of mesophyll chloroplasts, was 
mounted in a metal holder so that its 
abaxial surface faced the incident light 
beam, and was then immersed in liquid 
N 2 . The emission spectrum (Fig. 9) 
shows peaks near 686 and 740 nm, 
peaks usually attributed to pigments 
from PS II and PS I, respectively 
(Satoh and Butler, 1978). The spectrum 
from a suspension of chloroplast iso- 



— y> _ 




/ \ 

GUARD / \ 
CELLS / \ 

1/ \\ 


/ '> \ 

/ / \ \ 
It \ \ 
/ ' \ 
/ ' x 

II \ 
ij \ 


/ ' x 

T^-\ i I 


1 ,"*'!' 


660 700 740 780 


Fig. 9. The 77 K fluorescence emission spec- 
tra of chloroplasts from C. comosum obtained 
with a fluorospectrophotometer. Solid line, al- 
bino portion of a variegated leaf; dashes, isolated 
mesophyll chloroplasts from a green portion of 
the same leaf; dots and dashes, spectrum from 
an albino portion devoid of the abaxial epidermis. 

lated from a green portion of the same 
leaf shows identical peaks (Fig. 9). We 
suggest that guard cell chloroplasts of 
Chlorophytum have pigments of both 
PS I and PS II. 

Variable Fluorescence of Guard Cell 
Chloroplasts. The fluorescence induc- 
tion curve of higher plant chloroplasts 
reflects the light-dependent transition 
of PS II from a weakly fluorescent con- 
dition (F ) when all photochemical cen- 
ters of PS II are open and capable of 
performing a charge separation, to a 
more strongly fluorescence state (F m ) 
as these centers become closed. In 
terms of the quencher theory (Duysens 
and Sweers, 1963), the fluorescence rise 
is the consequence of photoreduction of 
the primary electron acceptor Q of 
PS II to its Q" forms (Melis and 
Duysens, 1979; Van Gorkom, 1974). 
According to the linear electron trans- 
port scheme, electrons accumulate on 
Q after the reduction of the plasto- 
quinone pool and the other electron ac- 
ceptors by PS II. Consequently, in 
dark-adapted chloroplasts, the kinetics 
of the fluorescence induction provide 
an indicator of the accumulation of 
electrons in a number of carriers lo- 
cated between the two photosystems. 
Figure 10a shows the light-induced flu- 
orescence rise curve in dark-adapted 

isolated mesophyll chloroplasts and 
guard cells from C. comosum. Both 
mesophyll and guard cell chloroplasts 
showed the typical kinetic transition 
from F to F m through the in- 
termediary fluorescence level (F p! ) (For- 
bush and Kok, 1968; Kautsky et al, 
1960; Malkin and Kok, 1966). The two 
time courses were similar, suggesting 
that PS II mediated the reduction of a 
pool of electron acceptors in both 
cases. With both preparations, irradia- 
tion at 700 nm was capable of restoring 
the fluorescence levels to their dark- 





~7 pl 




, 1,1,1,1 

1 2 3 4 5 

TIME (s) 

1.0 - 

0.8 - 

0.6 - 




- CELLS/''" 

/ / 


-/ / 

| + DCMU | 

r .. i l i 



0.2 03 
TIME (s) 



Fig. 10. Kinetics of the variable fluorescence 
yield from isolated mesophyll and guard cell 
chloroplasts from C. comosum in the absence of 
any artificial cofactors (a) and in the presence of 
the electron-transport inhibitor DCMU (b). Note 
the different time scales. The kinetic data have 
been normalized to the values of the initial fluo- 
rescence yield F (F = 0) and the maximum 
fluorescence yield F m (F m = 1.0). F pl , DCMU- 
insensitive intermediary fluorescence yield level. 



adapted state (data not shown), indi- 
cating that PS I was acting as a sink 
for electrons accumulated by PS II. 

The presence of a functional PS II in 
guard cell chloroplasts is further dem- 
onstrated by measurements of the vari- 
able fluorescence induction kinetics in 
the presence of the PS II inhibitor 
DCMU. Inasmuch as DCMU prevents 
the reoxidation of Q~ by the plasto- 
quinone pool (Duysens and Sweers, 
1963), fluorescence induction in its 
presence occurs in a much shorter pe- 
riod of time. Figure 10b shows that 
chloroplasts from both the mesophyll 
and guard cells of C. comosum respond 
to the inhibitor as predicted. Under 
these experimental conditions, the var- 
iable fluorescence induction reflects the 
reduction of the primary electron ac- 
ceptor of PS II only. 

These data indicate that guard cell 
chloroplasts have a functional PS II 
connected through an intermediate 
plastoquinone pool to PS I. A com- 
parison of the areas confined by the or- 
dinate, the F m level, and the induction 
curve, in the presence and absence of 
DCMU, provides an estimate of the rel- 
ative size of the plastoquinone pool 
(Thorne and Boardman, 1971). Our es- 

timate (data not shown) indicate that 
the pool sizes in both types of chloro- 
plasts are similar. These results are 
discussed more fully elsewhere (Zeiger 
etai, 1981). 


Duysens, L. N. M., and H. Sweers, in Studies on 
Microalgae and Photosynthetic Bacteria, 
353-372, University of Tokyo Press, Tokyo, 

Forbush, B., and B. Kok, Biochim. Biophys. 
Acta, 162, 243-253, 1968. 

Hsiao, T. C, in Encyclopedia of Plant Physiol- 
ogy, New Series, Vol. 2, V. Luttge and M. G. 
Pitman, eds., Springer, Berlin, 1976. 

Kautsky, H., W. Appel, and H. Amann, Bio- 
chem. Zeit., 322, 277-292, 1960. 

Malkin, S., and B. Kok, Biochim. Biophys. Acta, 
126, 413-432, 1966. 

Meidner, H., and T. A. Mansfield, Physiology of 
Stomata, McGraw-Hill, London, 1968. 

Melis, A., and L. N. M. Duysens, Photochem. 
and Photobiol, 29, 373-382, 1979. 

Satoh, K., and W. L. Butler, Biochim. Biophys. 
Acta, 502, 103-110, 1978. 

Thorne, S. W., and N. K. Boardman, Biochim. 
Biophys. Acta, 234, 113-125, 1971. 

Van Gorkom, H. J., Biochim. Biophys. Acta, 
347, 439-442, 1974. 

Zeiger, E., P. Armond, and A. Melis, Plant Phy- 
siol, 67, 17-20, 1981. 

Zelitch, I., Annu. Rev. Plant Physiol, 20, 
329-350, 1969. 




Anastasios Melis and Eduardo Zeiger 

High-resolution chlorophyll a fluo- 
rescence was employed in the study of 
kinetic transients from mesophyll and 
guard cell chloroplasts of Chlorophy- 
tum comosum. Like their mesophyll 
counterparts, guard cell chloroplasts 
showed the OPS transient (Papageor- 
giou, 1975), a manifestation of linear 
electron transport (Bannister and Rice, 
1968), and the slower MT transition re- 
flecting chloroplast photophosphoryla- 
tion. The MT transition of guard cell 
chloroplasts was sensitive to C0 2 , pro- 

viding the first evidence for a specific 
response of the guard cell chloroplast 
to CO 2 . 


Fluorescence Transients of Meso- 
phyll and Guard Cell Chloroplasts from 
Intact Tissue. Figure 11 compares the 
in vivo chlorophyll fluorescence tran- 
sients from dark-adapted mesophyll 
and guard cell chloroplasts. Mesophyll 
chloroplasts showed the typical tran- 



15 20 

TIME (s) 

Fig. 11. Time course of chlorophyll a fluores- 
cence from dark-adapted mesophyll and guard 
cell chloroplasts of Chlorophytum comosum in- 
tact leaf segments. The onset of actinic illumina- 
tion occurred at zero time. 

sients OPSMT, as reported previously 
(Govindjee and Papageorgiou, 1971). 
Guard cell chloroplasts showed a single 
transient that peaked at about the 
same time as P in mesophyll chloro- 
plasts, and a subsequent gradual decay 
to a steady-state level, which resem- 
bled the MT transition. 

We determined that the single fluo- 
rescence transient peak from guard cell 
chloroplasts responds like P in meso- 
phyll chloroplasts both in terms of its 
time of appearance under actinic illu- 
mination and its restoration kinetics in 
darkness (not shown). Guard cell chlo- 
roplasts also exhibit the MT transi- 
tion, as evidenced by the reversible 
slow fluorescence yield decrease occur- 
ring after 10 s of actinic illumination 
(Fig. 11). The SM transition, however, 
is either lacking in guard cell chloro- 
plasts or, if present, it is of small ampli- 

tude and is thus masked by other fea- 
tures of the fluorescence changes. 

Effect of C0 2 on the Fluorescence 
Transients from Guard Cell Chloro- 
plasts. The fluorescence transients 
from guard cell chloroplasts exposed to 
bicarbonate or Ca(OH) 2 were markedly 
different from their control. Figure 12 
shows the reversible effects of a 
CA(OH) 2 and bicarbonate solution on 
the fluorescence kinetics of guard cell 
chloroplasts. The single fluorescence 
transient peak GC of these chloro- 
plasts was largely unaffected by either 
treatment; however, the secondary flu- 
orescence yield decrease corresponding 
to the MT transition was significantly 
accelerated by Ca(OH) 2 or prevented 
by bicarbonate, a mode of action iden- 
tical to that in mesophyll chloroplasts. 

In the preceding article, (Zeiger et al, 
this Report), evidence was presented 
that guard cell chloroplasts operate a 
photosystem I- and II-dependent lin- 
ear electron transport chain. The pres- 
ent work provides evidence that guard 
cell chloroplasts show the fluorescence 
transient OPS, thus increasing the 
likelihood that guard cell chloroplasts 
generate reducing equivalents in the 
form of NADPH as a direct result of 
the electron transport activity in these 

The slow fluorescence decrease (MT 
transition) was also detected in guard 
cell chloroplast samples. In mesophyll 
chloroplasts, the MT transition has 
been correlated with the accumulation 
of high-energy compounds (ATP) and 
their intermediates (ApH, Aions) in the 
chloroplast stroma (Bennet et al, 1980; 
Briantais et al, 1979; Horton and 
Black, 1980; Krause, 1974). It appears, 
therefore, that guard cell chloroplasts 
have the capability of photophosphory- 
lation. The generation of metabolic 
energy in the form of ATP and possibly 
NADPH by the guard cell chloroplasts 
is of particular importance for stoma- 
tal physiology and strengthens the 
notion that these organelles play a cen- 
tral role in stomatal movements. 




10 15 20 

TIME (s) 

Fig. 12. Effect of Ca(OH) 2 and bicarbonate on the slow chlorophyll fluorescence decrease of 
guard cell chloroplasts. The same sample was illuminated three times with a 6-min dark interval be- 
tween the individual kinetic runs. The water surrounding the control sample was replaced by a 5 mM 
Ca(0H) 2 solution during the first dark interval. Similarly, the Ca(OH) 2 solution was replaced by a 
50-mM NaHC0 3 , pH 6.7 solution during the second dark incubation. Note the reversibility of the ef- 

The inhibition of the MT fluores- 
cence transition by C0 2 provides the 
first evidence for a specific response of 
the guard cell chloroplasts to C0 2 . 
Such a response was unexpected be- 
cause guard cell chloroplasts are re- 
portedly unable to fix C0 2 photosyn- 
thetically (Raschke, 1979; Outlaw et 
al, 1979). By analogy to mesophyll 
chloroplasts, C0 2 in guard cells must 
trigger photosynthetic energy consump- 
tion by an as yet unidentified process 
which prevents the formation of the 
high-energy state and the concomitant 
MT fluorescence transition. 


Bannister, T. T., and G. Rice, Biochim. Biophys. 
Acta, 162, 555-580, 1968. 

Bennet, J., K. E. Steinback, and C. J. Arntzen, 
Proc. Nat. Acad. Sci. USA, 77, 5253-5257, 

Briantais, J. M., C. Vernotte, M. Picaud, and 
G. H. Krause, Biochim. Biophys. Acta, 548, 
128-138, 1979. 

Govindjee, and G. Papageorgiou, in Photophysi- 
ology. Vol. 6, 1-46, A. C. Giese, ed., Academic 
Press, New York, 1971. 

Horton, P., and M. T. Black, FEBS Lett, 119, 
141-144, 1980. 

Krause, G. H., Biochim. Biophys. Acta, 292, 
715-728, 1974. 

Outlaw, W. H., J. Manchester, C. A. Dicarmelli, 
D. D. Randall, B. Rapp, and G. M. Vieth, Proc. 
Nat. Acad. Sci. USA, 76, 6371-6375, 1979. 

Papageorgiou, G., in Bioenergetics of Photosyn- 
thesis, 319-371, Govindjee, ed., Academic 
Press, New York, 1975. 

Raschke, K., in Encyclopedia of Plant Physi- 
ology, New Series, Vol. 7, 384-441, W. Haupt 
and M. E. Feinleib, eds., Springer, Berlin, 




RED ALGA Porphyra perforata, 

THE LIVERWORT Porella navicularis, AND THE 

ISOLATED LICHEN GREEN ALGA Trebouxia pyriformis 

Gunnar Oquist and David C. Fork 

We showed in Year Book 79 (193- 
197) that desiccation of the intertidal 
red alga Porphyra perforata caused a 
loss of photosystem II (PS II) fluores- 
cence emission (F) at 77 K. It was con- 
cluded from emission and excitation 
spectra analyses that this was achieved 
by an increased energy transfer from 
the phycobilins to PS I. Desiccation of 
mosses such as Dendroalsia abietinia 
and Scleropodium tourettei, of liver- 
worts such as Porella bolanderi and 
Porella navicularis, and of the lichen 
Cladonia implexa and the symbionic 
green alga Trebouxia pyriformis, iso- 
lated from Cladonia also induced a 
preferential loss of PS II fluorescence 
emission at 77 K. This change is shown 
in Fig. 13 for Porella navicularis and 
Trebouxia pyriformis. Light desicca- 
tion caused a stronger loss of PS II 
emission than did dark desiccation. 

The reason for the preferential loss of 
PS II emission upon desiccation was 
investigated by following the kinetics 
of PS I and PS II emission during 
PS II trap closure at 77 K (Butler, 
1978). Porphyra represents a species 
that can tolerate desiccation under sun 
exposure at low tides, whereas Porella 
navicularis is a species that undergoes 
desiccation in its very shaded habitat. 
Porella was collected from the base of a 
live oak tree trunk {Quercus agrifolia) 
in the redwood forested area of Wood- 
side, California. Trebouxia was obtain- 
ed from the culture collection of algae 
at the University of Texas at Austin, 
Texas (#1712). It was grown on agar 
slants containing the nutrient medium 
described elsewhere (Sigfridsson and 
Oquist, 1980). The microprocessor- 
based spectrofluorimeter described 

earlier ( Year Book 78, 196-199) was us- 
ed in these experiments. 

It has been demonstrated that Fj = 
f(Fjj) when the PS II traps are closed 
at 77 K (Kitajima and Butler, 1975). 
The function gives a straight line in an 
X-Y-plot, and it can be shown that the 
slope of the line 

Fj/F n = k 

Till- 1) 



where a, the fraction of light initially 
absorbed by PS I, is proportional to 
the intercept obtained when the straight 
line is extrapolated back to the Y-axis, 
which represents the PS I emission 
(see Fork, Oquist, and Powles, this Re- 

Figures 14, 15, and 16 show such 
X-Y plots for the PS II and PS I emis- 
sion of Porphyra (collected in Septem- 
ber because spring-collected algae 
[Year Book 79, 193-197] showed in the 
dry state no PS II emission at all), 
Porella, and Trebouxia, respectively. 
F and F m values are summarized in 
Table 5. Desiccation of Porella caused 
a very marked decrease of both F and 
F m at 685 nm (PS II) and 730 nm 
(PS I). Trebouxia responded with only 
minor decreases whereas Porphyra was 
intermediate. In Porphyra, desiccation 
induced a small increase of the slope of 
the straight-line plot (10%) and a in- 
creased by 47 units (Fig. 14). For rea- 
sons we do not fully understand, the 
plot of wet Porphyra always crossed 
the Y-axis below zero. It means, how- 
ever, that PS II of wet Porphyra re- 
ceives some excitation energy which 
does not transfer to PS I. This fraction 
should be proportional to the intercept 
on the PS II axis when the PS I emis- 





Wavelength, nm 

Fig. 13. Low-temperature (77 K) fluorescence emission spectra with deconvoluted emission 
bands of wet, dark-dried, and light-dried Porelia navicularis and Trebouxia piriformis. Fluorescence 
was sensitized by chlorophyll b excitation (478 nm). The emission was corrected to equal absorp- 
tance at 478 nm, as indicated in the second footnote of Table 5. The spectra were matched to equal 
heights of the dominant peaks using the multiplication factor given. The fluorescence properties of 
Porelia represent equal leaf areas and those of Trebouxia equal chlorophyll concentrations. The error 
curves show the differences between the recorded (dotted-curve) spectrum and the sum of the com- 

sion is zero. The slope of the straight- 
line plot increased for Porelia from 0.68 
to 1.39 upon desiccation (Fig. 15), 
whereas for Trebouxia (Fig. 16) the 
slope only increased from 0.80 to 0.85. 
The possible effects on the slopes of 
the straight-line plots by desiccation- 
induced changes in k r{II ^ }) , k FlI , and <j>Fj 
were evaluated. The procedure for do- 
ing this is exemplified for Porphyra. 
The effect of desiccation on 4>F { was 

estimated from values of F730 when 
exciting chlorophyll a at 433 nm (Table 
5); 0F/ dry = 0.65 • 0F 7 wet. The 
straight-line plot, dry (corr </> 7 ), of Fig. 
14 was obtained by dividing the F730 
values of dry Porphyra by 0.65 in order 
to correct for the desiccation-induced 
drop of <t>Fj. The estimation of possible 
desiccation-induced effects on k F[I is 
less straightforward, but it was done 
as follows: First, F o 730 of dry Por- 




"30 i 

dry (corr 0Fj) 


100 200 

Fluorescence 695 nm, rel. 

Fig. 14. F695 (PS II) vs. F730 (PS I) of wet, 
dry, and dry (corr <t>F { ) Porphyra perforata. See 
text for the method used to obtain the line for 
dry (corr <j>Fj). Excitation was on phycoerythrin 
using 554-nm radiation. Correction to equal ab- 
sorptance at 554 nm was done according to the 
second footnote in Table 5. The equation of the 
straight-line plots are indicated, and the linear 
regression coefficients are: wet, r 2 — 0.99; dry, 
r 2 = 0.82; dry (corr <t>Fj), r 2 = 0.82. The averages 
of three experiments are shown. 

phyra was corrected to what it would 
have been if desiccation had not af- 
fected 4>Fj. This correction increased 
F o 730 from 99 to 153 (Table 5). Second, 
the 1.6-fold increase in F o 730 upon dry- 
ing seen after the above correction had 
been applied was concluded to be 
caused by an increase in a and/or 
^■T(ii-i)(o)- ^ we assume therefore that 
F 695 was decreased with the same 
factor that caused the increase in 
F o 730 (because increases in a and 
k T m~i)( ) would lower F 695), we can 
calculate that F 695 would be lowered 
from 93 to 57 upon desiccation. As the 
calculated value of F n 695 is identical 





dry (corr 0F,) 

ft" 1 "' 


/# dry 




50 100 

Fluorescence 685 nm. rel. 

Fig. 15. F685 (PS II) vs. F730 (PS I) of wet, 
dry, and dry (corr <t>Fj) Porella navicularis. The 
dry (corr 4>Fj) values were obtained as explained 
in the text. The equations of the straight lines 
are indicated, and the linear regression coeffi- 
cients were: wet, r 2 = 0.93; dry, r 2 = 0.92; dry 
(corr c{>Fj), r 2 = 0.91. The averages of three ex- 
periments are given. 

with the measured one (Table 5), we as- 
sume that desiccation did not affect 
k FlI and that the slopes of the straight 
lines (Fig. 14, wet and dry [corr <j)Fj\\ 
would be proportional to k T{II ^jy i.e., 
krai-i) increased 1.7 times and a in- 
creased 55 units upon desiccation. A 
similar calculation performed on dark- 
dried Porella revealed that the prefer- 
ential loss of PS II emission must be 
caused first of all by a strong decrease 
in kFu (^T(//-/))' ano ^ a even decreased 
slightly on desiccation. In dark-dried 
Trebouxia, k FlI was not significantly 
affected, and the increase of the slope 
of the line dry (corr (j>Fj) must be the 
result of a 10-20% increase of k T{II ^j). 
Whether or not k D]I (radiationless de- 
cay in PS II) was affected by desicca- 
tion cannot be evaluated. The more pro- 
nounced desiccation-induced decreases 
in F m 685 than in F 685 in both Porella 
and Trebouxia must be the result of a 
specific quenching of F v , possibly by 
some quenching by P680 + Q~ which 
may accumulate to some extent in dry 




100 - 

Fluorescence 685 nm, rel. 

Fig. 16. .F685 (PS II) of wet, dry, and dry (corr <t>Fj) Trebouxia pyriformis. The dry (corr 4>Fj) 
values were obtained as explained in the text. The equations of the straight lines are indicated, and 
the linear regression coefficients were: wet, r 2 — 0.95; dry, r 2 = 0.98; dry (corr (frFj), r 2 = 0.98. The 
averages of three experiments are given. 

samples exposed to light at 77 K. Light 
desiccations induced a stronger loss of 
PS II emission than did dark desicca- 
tion (Fig. 13). Relative two-point action 
spectra of Porella and Trebouxia 
(chlorophyll a and chlorophyll b excita- 
tion with 433 nm and 478 nm, respec- 
tively) showed that light merely accen- 
tuates the changes taking place in 
&r(//-/)> k FlI , and a in dark-dried 

In conclusion, the preferential loss of 
PS II fluorescence emission at 77 °C in 
Porphyra was mediated by increases in 
krai-i) and a, whereas in Porella it was 
caused preferentially by a decrease in 
k FlI . The small loss of PS II emission 
upon desiccation of Trebouxia was the 
consequence of a small increase in 
h-Tai-iy whereas k FlI and a were unaf- 

Ecological implications can be sug- 
gested for the differences observed in 
energy distribution in the plants stud- 
ied. Since Porphyra can be desiccated 
at low tide and be exposed to bright 
sunlight, a mechanism has been devel- 

oped to transfer excess excitation en- 
ergy to PS I where it can be trans- 
formed eventually into heat. Such a 
mechanism would protect PS II from 
photodynamic damage in the desic- 
cated state (compare Sigfridsson and 
Oquist, 1980). Porella, on the other 
hand, is found in extremely shaded 
habitats and is not exposed to bright 
sun, either in the wet or dry state. Per- 
haps PS II of Porella can handle small 
amounts of excess excitation in the 
desiccated state without damage to 
PS II. Porella has therefore not devel- 
oped a specific mechanism for coping 
with large amounts of excess excita- 
tion energy. It is interesting to note in 
this regard that the leaves of Porella 
roll together in the dry state and thus 
decrease their absorption of light. 


The authors are grateful to Professor W. B. 
Schofield of the University of British Columbia, 
who provided identification of the mosses and 
liverworts used in this work. 


TABLE 5. The F and F m Values of Wet and Dry Porphyra, Porella, and Trebouxia.* 

433 nm 

554 nm 




F m 695 

F o 730 

F m 730 

Porphyra, wet 
Porphyra, dark-dried 
Porphyra, dark-dried 
(corr <t>Fj) 

88 ± 4 
57 ± 2 + 

93 ± 5 

57 ± 9++ 

200 ± 21 
110 ± 10++ 

94 ± 3 
99 ± 11++ 

153 § 

237 ± 2 
185 ± 19++' 

284 § 

700 nm 

478 nm 



F 685 

F m 685 

F o 730 

F m 730 

Porella, wet 
Porella, dark-dried 
Porella, dark-dried 
(corr 4>Fj) 

124 ± 6 
64 ± 9+ 

47 ± 2 
13 ± 1++ 

114 ± 2 

21 ± 1++ 

66 ± 4 

25 ± 1++ 

48 ± 2 § 

112 ± 6 

35 ± 3++ 

68 ± 6 § 

700 nm 

478 nm 



F 685 

F m 685 

F 715 

F m 715 

Trebouxia, wet 
Trebouxia, dark-dried 
Trebouxia, dark-dried 
(corr <t>Fj) 

73 + 

63 ± 6 

58 ± 5++ 

180 ± 18 
113 ± 9 ++ 

63 ± 8 
53 ± 1++ 

64 ± 1 § 

154 ± 23 
102 ± 12++ 

122 ± 13 § 

*In Porphyra, F695 (PS II) and F730 (PS I) were sensitized by chlorophyll (433-nm) or 
phycoerythrin (554-nm) excitation. In Porella and Trebouxia, F685 (PS II) and F730/F715 (PS I) 
were sensitized by chlorophyll a (700-nm) or chlorophyll 6 (478-nm) excitation. Standard deviations 
for n = 3-4 are given. 

+F730 (Porphyra) was corrected to what it would have been if the dry sample had the same absorp- 
tance as the three-layered wet sample at 433 nm. The measured value was multiplied by 1.15. The 
F715 {Trebouxia) and F730 (Porella) values were corrected as if the dry and wet samples had equal 
absorptance at 700 nm. For this correction the measured values of dry Porella were multiplied by 1.0 
and those of dry Trebouxia by 0.68. 

++To give the dry samples the same absorptance at 554 nm as three layers of wet Porphyra, the 
measured fluorescence values of dry Porphyra were multiplied by 1.24. To give the dry samples the 
same absorptance as the wet ones at 478 nm, the values of dry Porella and Trebouxia were multiplied 
by 0.92 and 0.89, respectively. 

§In addition to the correction listed above (ft), these fluorescence values have been corrected to 
what they would have been if $F 7 was unaffected by desiccation; for Porphyra <t>Fj dry = 0.65 4>Fj 
wet, for Porella 4>Fj dry = 0.52 0F 7 wet, and for Trebouxia <$>F[ dry = 0.84 cj>Fj wet (see text). 


Butler, W. L., Annu. Rev. Plant Physiol., 
345-378, 1978. 


Kitajima, M., and W. L. Butler, Biochim. Bio- 

phys. Acta, 408, 297-305, 1975. 
Sigfridsson, G., and G. Oquist, Physiol. Plant, 

49, 329-335, 1980. 



RED ALGA Porphyra perforata 

David C. Fork and Gunnar Oquist 

We showed previously {Year Book 
79, 193-197) that desiccation of the 
red alga Porphyra perforata (to 91% 
water loss) caused increased transfer of 
quanta absorbed by PS II to PS I. It 
was suggested that this effect may pro- 
tect Porphyra from photodynamic dam- 
to PS II when the plant is exposed to 
full sunlight and extreme desiccation 
during low-tide periods. The former 
study used measurements of fluores- 
cence of PS I (730 nm) and PS II (685 
nm and 697 nm) made at 77 K, and the 
model of Butler (1978), to suggest how 
desiccation affects energy distribution 
between the two photosystems of 

Since Porphyra is unusual in having 
a fluorescence emission band at room 
temperature that is sensitized by PS I 
(Fork, Oquist, and Hoch, this Report), 
we wanted to take advantage of this 
property to extend our investigation of 
how desiccation affects energy transfer 
between two photosystems of photo- 
synthesis at physiological tempera- 

The fluorescence emission and exci- 
tation spectra were measured as de- 
scribed in another section of this Re- 
port (Oquist and Fork; Fork, Oquist, and 
Hoch). The kinetics of fluorescence 
were measured either with blue or 
green actinic light, the intensity of 
which was decreased until fluorescence 
could be measured with a two-pen 
chart recorder. Porphyra, which is only 
one cell layer in thickness, could be 
easily dried in darkness by passing an 
air stream over pieces of the thalli. 
Drying was also done in the presence of 
strong white light. 

Excitation of Porphyra with blue 
light that excites PS I of photosyn- 

thesis predominantly gives rise to the 
emission spectrum shown in Fig. 17, 
curve A. As was discussed in another 
section of this Report (Fork, Oquist, 
and Hoch), PS I excitation in Porphyra 
produced a distinct long wavelength 
fluorescence emission band at room 
temperature. Figure 17, curve B, shows 
that desiccation in the light to 91% 

1 ' 


1 1 1 1 1 l l l l 1 1 1 1 1 




/ V*C, Dark dried _ 


1 1 1 
//' \ 1 

// 4— B. Light dried 

// M — 

I M 

// M 
// M 

J A 


727 \\ 

1 i \\ 

A, Wet — \ V 0V 


1 i 

l . , 

i i i i 1 i i i i 1 i i i i 


700 750 800 

Wavelength, nm. 


Fig. 17. Fluorescence emission spectra mea- 
sured in wet and dry Porphyra perforata at 24°C 
using chlorophyll a excitation (433-nm interfer- 
ence filter, 13-nm half-bandwidth; Corning glass 
filter 5-60; and Bakers Calflex C, 2 ^mol quanta 
m~ 2 s" 1 ). (A) Wet, dark-adapted sample. (B) 
Dried in strong, white light (2,100 quanta m~ 2 
s _1 ). (C) Sample dried in darkness. Three layers 
of the wet and one layer of the unstretched dried 
alga were used in order to have samples with ap- 
proximately equal optical properties. In all ex- 
periments the spectra were measured with the 
same photomultiplier voltage and gain setting. 



water loss produced an increase in the 
intrinsic fluorescence yield of PS I (of 
3.6 times) and a 10-nm shift of the 
727-nm peak to longer wavelengths. 
When the Porphyra was dried in 
darkness, the fluorescence yield in- 
creased 4.3 times (Fig. 17, curve C). 

Excitation of PS II in Porphyra with 
green light produced the spectrum 
shown in Fig. 18, curve A. The 
shoulder near 643 nm and the peak at 
660 nm correspond to emissions from 
phycocyanin and allophycocyanin, 
respectively. Chlorophyll a of PS II 
produces the 687-nm peak and, as 
discussed previously, PS I produces 
the long wavelength shoulder at 727 
nm. Curve A of Fig. 18 represents the 
spectrum for fluorescence in the F Q 
state, since Q, the primary electron ac- 
ceptor of PS II, remained largely in the 

6 - 

4 — 


1 1 ' 

i i i 

i i i i I i i i 









\-B, Light dried 





/ 660\ 


;<687 727 

V. |_ 

A, Wet-" 











-, 1 . , 

i 1 i 

i i i 

I i i i i i i "n 


2 — 

650 700 750 800 850 

Wavelength, nm. 

Fig. 18. Fluorescence emission spectra mea- 
sured at 24°C in wet and dry Porphyra perforata 
using phycoerythrin excitation (543-nm inter- 
ference filter, 9-nm half-bandwidth. Corning 
glass filter 4-96; Calflex C, 9 //mol quanta m -2 
s _1 ). (A) Wet, dark-adapted (three layers). (B) 
Dried in strong, white light (one layer). 

oxidized state in the relatively weak ac- 
tinic light intensities used. If we 
treated Porphyra with DCMU, then Q 
became reduced as a result of the inter- 
ruption of normal electron flow be- 
tween the two photosystems. Under 
this condition (F m ) we saw increases of 
3.6 times at both the 687-nm and 
730-nm bands compared to the F spec- 
trum (Fork, Oquist, and Hoch, this 
Report). DCMU, therefore, increased 
fluorescence yield but did not change 
the shape of the fluorescence emission 
spectrum. By contrast, desiccation 
produced a large change in the shape of 
the emission spectrum (Fig. 18, curve 
B). In this case, Porphyra dried in light 
showed a large increase of the far-red 
emission band, a loss of the 687-nm 
band, and an increase of the allophyco- 
cyanin band at 658 nm. As was seen by 
measurements of fluorescence kinetics 
in the light-dried alga, the reaction 
center of PS II (Q) does not become 
reoxidized in darkness. 

Since Fig. 17 shows that the intrinsic 
fluorescence yield increased upon des- 
iccation of Porphyra either in the light 
or the dark, we wanted to replot the re- 
sults of the phycoerythrin-sensitized 
spectra so that this drying-induced in- 
trinsic fluorescence yield increase of 
PS I would be removed. After doing 
this manipulation, it would be possible 
to see more clearly if drying produced 
effects on emission spectra other than 
the intrinsic fluorescence yield in- 
creases of PS I. In Fig. 19, the relative 
fluorescence emission spectra mea- 
sured using green actinic light were 
corrected for equal intrinsic fluores- 
cence yields of PS I for both dark-dried 
and light-dried Porphyra (Fig. 19, 
curves C and B) and were then com- 
pared to a wet sample (Fig. 19, curve 
A). These spectra show that desicca- 
tion produced dramatic increases in the 
relative fluorescence yields at 737 nm 
(PS I) and decreases of fluorescence at 
687 nm (PS II). In Porphyra dried in 
strong light (Fig. 19, curve B), there 
was almost a complete disappearance 



— C, Dark dried 

B.Light dried 


_i_i i i i i_ 

_1_1 I I L_ 


700 750 800 

Wavelength, nm. 


Fig. 19. Fluorescence emission spectra mea- 
sured at 24 °C in wet and dry Porphyra perforata 
using phycoerythrin excitation, as for Fig. 18. 
(A) Wet (three layers). (B) Dried in strong, white 
light and corrected for fluorescence yield in- 
crease of chlorophyll a upon drying, as explained 
in the text. (C) Dried in darkness and corrected 
as described in (B). 

of the PS II 687-nm band. In Porphyra 
desiccated in darkness, only a small 
shoulder at 687 nm remained (Fig. 19, 
curve C). It is interesting to note that, 
except for the 658-nm allophycocyanin 
band, desiccation in strong light pro- 
duced less fluorescence yield increase 
at 737 nm than did desiccation in the 
dark. The 737-nm and 687-nm fluores- 
cence bands were decreased by about 
equal proportions upon desiccation in 
strong light, as compared to the dark- 
dried sample. 

We measured the kinetics at 24 °C of 
the rise of 730-nm fluorescence upon 
excitation of wet and desiccated Por- 
phyra with green actinic light. Since 
Porphyra is photosynthetically active 
at 24 °C, the value of F Q was the same 
as F m because Q did not accumulate in 
the reduced form under the relatively 

low light intensities used. Adding 
DCMU to the wet sample produced a 
fluorescence increase of 2.5 times at 
730 and a similar increase at 687 nm 
(Fork, Oquist, and Hoch, this Report). 
Illumination of a sample dried in the 
dark with green light produced a 1.4- 
fold increase of fluorescence. This sug- 
gests that PS II traps could still be 
reduced to a substantial degree (56%) 
in the dry state. Light-dried Porphyra, 
on the other hand, did not have the 
ability to reduce PS II traps; both F 
and F m were the same, probably 
because all PS II traps were closed in 
the light-dried sample. 

The F m value of fluorescence at 730 
nm in dark-dried Porphyra was double 
that in wet Porphyra (no DCMU) and 
1.4 times higher than that of light- 
dried Porphyra. This finding agrees 
quite well with emission spectra re- 
sults presented in Fig. 19. The desicca- 
tion-induced increase of the fluores- 
cence yield at 730 nm and the decrease 
at 687 nm seen in emission spectra and 
kinetic measurements may arise from 
an enhanced energy transfer from 
PS II to PS I as the result of drying. If 
this is the case, then the excitation 
spectrum of dry Porphyra would show 
an enhanced PS II (phycobilin) effec- 
tiveness to excite 730-nm fluorescence 
in the dry samples, as compared to the 

Figure 20 shows excitation spectra 
for wet and dry Porphyra. The relative 
effectiveness of green light (548, 570 
nm) absorbed by the phycobilins to ex- 
cite the far-red fluorescence band at 
730 nm is higher in both the light-dried 
and dark-dried Porphyra than in the 
wet alga. 

The changes of the fluorescence prop- 
erties of Porphyra dried in the dark, 
such as the reduced efficiency for PS II 
trap closure, the almost complete loss 
of PS II emission at 687 nm with con- 
comitant relative increase of PS I emis- 
sion at 730 nm, and the increased phy- 
coerythrin/chlorophyll efficiency ratio 
for sensitizing PS I emission, are quali- 



1 ' ' 

i | i i i i 1 i i i i | i i i i ■ 






/ \ 














— 420 


■ / \ 

Light dried / / X^ 

445 5C f/ / Wet-\^ 
r\ Dark dried \J? / 

1 v> /s 


l~i i i 

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 

400 450 500 550 

Wavelength, nm. 


Fig. 20. Excitation spectra measured at 
room temperature for 730-nm fluorescence emis- 
sion in wet Porphyra perforata and in Porphyra 
dried in strong light and in the dark. 

tatively similar to the fluorescence 
changes observed at 77 K (Year Book 
79, 193-197; Oquist and Fork, 1981). It 
can be concluded that Porphyra, even 
at room temperature, responds to des- 
iccation by producing a strong increase 
in the distribution of excitation energy 
from phycoerythrin to PS I. Whether 
this shift is mediated by increases in a 
or spillover (see Butler, 1978) cannot be 
determined from the data presented, 
since there was almost a complete ab- 
sence of the PS II emission at 685 nm 
in Porphyra desiccated at room temper- 
ature. However, the observation that 
the F value of the dark-dried and the 
F m of the wet Porphyra ( + DCMU) 
were about the same favors the as- 
sumption that the desiccation-induced 
increase in energy transfer from phyco- 
erythrin to PS I at room temperature 
is mediated by a substantial increase in 

a. A similar interpretation was made 
for these algae studied at 77 K (Oquist 
and Fork, 1981). The loss of the capac- 
ity for trap closure by about one-half, 
induced by desiccation, can also be ex- 
plained by an a increase. It is possible, 
however, that reaction center quench- 
ing may account for some of this loss in 
the dark-dried sample (see below). 

Measurements of the kinetics of flu- 
orescence of Porphyra dried in strong 
light showed that F was not much af- 
fected, and part of F v was decreased in 
comparison with dark-dried algae. This 
F v loss was reflected in the emission 
spectrum shown in Fig. 19, curve B, 
where, as noted previously, the relative 
fluorescence yield was lower in the 
light-dried sample at both 687 and 737 
nm. Both the 687-nm and 737-nm bands 
were decreased proportionately. Loss 
of F v but not F suggests a mechanism 
of reaction center quenching (Butler, 
1978). This type of quenching occurs 
under conditions where PS II reaction 
centers accumulate in the state where 
P680 is oxidized and Q is reduced (P680 + 
Q~). If the decreased fluorescence in 
the light-dried sample compared to the 
dark-dried sample is caused by an in- 
crease in the rate constant for nonradi- 
ative decay of fluorescence (k d ), then 
decreases at both F and F m would 
have been seen. It is possible that the 
decreased capacity for trap closure of 
about 50% in the dark-dried Porphyra 
may have been caused to some degree 
by this type of reaction center quench- 

Our previous conclusion that desic- 
cation gives rise to an increased energy 
transfer from PS II to PS I was based 
on studies made at liquid nitrogen tem- 
peratures (Year Booh 79, 193-197). The 
present studies at physiological tem- 
peratures also favor the interpretation 
of increased energy transfer to PS I as 
a result of desiccation. As we noted pre- 
viously, this mechanism may serve to 
protect the more light-sensitive reac- 
tion centers of PS II from photochemi- 
cal damage when these desiccation-tol- 



erant algae are exposed during periods Reference 

Of low tide to both extreme drying and Butler, W. L., Annu. Rev. Plant Physiol, 29, 

high light intensities. 345-378, 1978. 



RED ALGA Porphyra perforata 

David C. Fork, Gunnar Oquist, and George Hoch 

At 77 K a fluorescence band at 725 
nm can be seen clearly in the red and 
blue-green algae, but at room tempera- 
ture this band is usually not seen (Goed- 
heer, 1972). However, Murata and Taka- 
miya (1967) saw a peak in this region in 
the room-temperature fluorescence emis- 
sion spectrum that was sensitized by 
blue light in the red alga Porphyra yez- 

A room-temperature fluorescence 
emission band that can be shown clearly 
to originate in PS I would be useful, 
since it would permit measurements at 
physiological temperatures of photosyn- 
thetic reactions that can be followed 
using fluorescence techniques (Butler, 
1978). We have seen a room-tempera- 
ture fluorescence band at 730 nm in 
Porphyra perforata that is emitted 
from PS I. 

Porphyra perforata used in these ex- 
periments was collected along the Pa- 
cific Coast north of Santa Cruz, Califor- 
nia, and kept under illumination in 
open dishes of seawater held at 13 °C. 
Fluorescence emission spectra were 
measured with the microprocessor- 
based system described previously 
{Year Book 79, 196-199). The alga, 
which is one cell layer thick, was held 
submerged in a Dewar at 24 °C. Excit- 
ing and fluorescent light were directed 
to and collected by a quartz light guide 
positioned just on the surface of the 
algal thallus. Emission spectra were 
corrected by using a standard lamp. 

Fluorescence excitation spectra were 
done with the same system except ac- 
tinic light from a monochromator was 

directed on the sample with the light 
guide. The fluorescence was conducted 
by another light guide to a Hamamatsu 
TV R928 photomultiplier fitted with 
either a 730- or 685-nm interference 
filter (5-nm half-bandwidth). Excita- 
tion spectra were corrected by a tech- 
nique that used Rhodamine B as a flu- 
orescence standard. 

Fluorescence kinetics were measured 
with a trifurcate fiber bundle. One 
branch served to illuminate the sample 
and the other two conducted fluores- 
cence at two different wavelengths to 
two photomultipliers having 685- and 
730-nm interference filters, respec- 
tively. The actinic light was kept low 
enough that the kinetics of fluorescence 
could be followed with a two-pen strip 
chart recorder. 

Porphyra illuminated at room tem- 
perature with green actinic light had 
an emission spectrum (Fig. 21, curve 
A) with three peaks at 659 nm (allophy- 
cocyanin), 687 nm (Chi a of PS II), and 
732 nm. As will be described, the 732-nm 
band is apparently emitted from Chi a 
of PS I. This spectrum represents the 
initial F fluorescence state, since it 
was measured in a physiologically ac- 
tive thallus under nonsaturating green 
actinic light. In the F state all the 
photochemical reaction centers are in 
the open (oxidized state). After addi- 
tion of DCMU to this sample, the emis- 
sion spectrum had maxima at 685 and 
734 nm and a shoulder near 663 nm (Fig. 
21, curve B). DCMU blocks electron 
transport between PS II and PS I, and 
reaction centers of PS II accumulate in 



o> 4 

i — rn — i — i — i — r~i — i — I — r 

Green actinic /"N, 

/ \ 

"i i i i i i r 


685 / 

A / 

/ \ / / V-i— c - Difference 
/' v / \ \ 

/ r ' * 'i 

650 700 750 800 850 

Wavelength, nm. 

Fig. 21. Room-temperature fluorescence emis- 
sion spectra of Porphyra perforata sensitized by 
phycobilin excitation (550 nm, half-bandwidth 5 
nm). The fluorescence analyzing monochromator 
had a half-bandwidth of 1.6 nm. In curve B, 
DCMU (10 fxM) was added to the same sample 
used for measurement of curve A. The same sen- 
sitivities were used for curves A and B. 

the reduced state. In this condition, 
fluorescence attains its maximum level 
(F m ), as all photochemical reaction cen- 
ters are reduced. The difference be- 
tween curves A and B of Fig. 21 (curve 
C) represents the fluorescence spec- 
trum having a variable yield (F v ). This 
spectrum has peaks at 685 and 734 nm 
and no phycobilin peaks, indicating 
that DCMU did not interrupt energy 
transfer from the phycobilins to chloro- 
phyll a fluorescing at 685 nm. The in- 
creased fluorescence intensity at 734 
nm after DCMU addition does not orig- 
inate in PS I but comes from energy 
absorbed by PS II chlorophyll and 
transferred to PS I because of the 
closed PS II reaction centers (Butler, 

Excitation of PS I of Porphyra by 
blue light gives a fluorescence emission 
spectrum with only one major peak at 

732 nm (and a minor inflection around 
687 nm) (Fig. 22, curve A). Adding 
DCMU increased both the 732- and 
685-nm fluorescence bands (Fig. 22, 
curve B). 

Excitation spectra (Fig. 23) show 
that blue light (440 nm) is more effi- 
cient in exciting fluorescence at 730 nm 
in Porphyra than is green actinic light. 
By contrast, green light at 553 and 568 
nm is more efficient than blue light in 
exciting fluorescence at 685 nm. These 
measurements represent excitation 
spectra for F fluorescence, since they 
were done at room temperature, in the 
absence of DCMU, and under rela- 
tively low light intensities. These ex- 
citation spectra, plus the fluorescence 
emission spectra described earlier, and 
the results of Murata (1967), demon- 
strate that the long-wavelength fluo- 
rescence band in Porphyra at room 
temperature originates from chloro- 
phyll associated with PS I. These re- 

o — 

Blue actinic 




700 750 

Wavelength, nm. 



Fig. 22. Room-temperature fluorescence emis- 
sion spectra of P. perforata sensitized by chloro- 
phyll a excitation (433 nm, half-bandwidth 12 
nm). Other conditions, as for Fig. 21. Curves A 
and B were measured with the same sensitivi- 

n — i — i — r 

6 - 

-i — rn — r 

~r~i — r— r 

I l l I 

\ f =730 


i i i i I i ' ' i I i i i i I i i 
400 450 500 550 

Wavelength, nm. 


suits suggest that in Porphyra it is 
possible at room temperature to mea- 
sure both PS I and PS II activities 
using fluorescence techniques. 


Butler, W. L., Annu. Rev. Plant Physiol., 29, 
345-378, 1978. 

Goedheer, J. C, Annu. Rev. Plant Physiol., 23, 
87-112, 1972. 

Murata, N., and A. Takamiya, Plant Cell Phy- 
siol, 8, 683-694, 1967. 


Fig. 23. Room-temperature fluorescence exci- 
tation spectra for 685 and 730 nm fluorescence 
emission in P. perforata. The half-bandwidths of 
the excitation wavelengths were 3.2 nm. 


David C. Fork and Gunnar Oquist 


Fluorescence yield changes have 
been suggested by Duysens and Sweers 
(1963) to reflect changes in the redox 
state of the primary electron acceptor 
(Q) of photosystem II of photosynthe- 
sis. The characteristic fluorescence 
yield changes seen upon reduction of Q 
can still be seen at the temperature of 
liquid nitrogen (77 K). Only primary 
photochemical reactions can function 
at 77 K; all other reactions are inhib- 
ited (Amesz, 1977). 

We have seen a biphasic dark reoxi- 
dation of Q in bean leaves at 77 K that 
is not affected by DCMU. This result 
may be a reflection of the heterogeneous 
nature of photosystem II. 

The bean leaves used in these experi- 
ments were grown in the greenhouse 
under natural illumination during late 
summer at 25 °C day and 20 °C night 
temperatures. The leaves were either 
used directly or vacuum infiltrated 
with DCMU. 

Fluorescence induction at 695 nm 
was followed with the apparatus de- 
scribed by Fork and Oquist (this Re- 
port). The leaves were dark adapted 
and cooled to 77 K and then illuminated 
with actinic light until no further fluo- 
rescence increases took place. After 
dark periods of varying lengths the ac- 
tinic light was again turned on and the 
fluorescence rise measured. 



Figure 24 shows the kinetics of reoxi- 
dation of reduced Q at 77 K in untreated 
bean leaves and in leaves infiltrated 
with DCMU. It can be seen that DCMU 
did not have any effect on the rate of Q 
reoxidation at liquid nitrogen tempera- 
ture. The curve of Fig. 24 could be de- 
convoluted into two first-order decay 
components, a fast component having a 
half-life of 0.8 min and a slow component 
having a half-life of 180 min. The fast 
component constitutes about 13% of the 
total fluorescence recovery seen at 77 K. 

Murata et al. (1973) saw dark re- 
covery of fluorescence induction at 77 K 
in isolated spinach chloroplasts that 
showed maximum recovery of the vari- 
able fluorescence after about 20 min of 
darkness. After that, fluorescence re- 
mained unchanged for 60 min of dark- 

Measurements by Melis and Homann 
(1975, 1976) of the fluorescence increase 

10 20 

Time, min 

Fig. 24. Kinetics of dark reoxidation of Q in a 
bean leaf at 77 K. Fluorescence was measured at 
695 nm (Infra Red Industries interference filter, 
5-nm half-bandwidth) with chlorophyll b excita- 
tion at 478 nm (0.5 ^mol quanta m~ 2 s _1 , Bal- 
zers interference filter 478 nm, half-bandwidth 9 
nm; Corning 5-60; Bakers Calflex C). The 
amount of Q reduced at varying dark times was 
taken as the log of the percent (F m — F )IF m . 
Open circles, no DCMU; solid circles, plus 4 X 
10~ 5 M DCMU. 

in chloroplasts treated with DCMU 
showed it to be biphasic. Two different 
kinetic components could be distin- 
guished, a fast (a) component and a 
slow (/3) component. These and other 
measurements (Melis and Duysens, 
1979; Rijgersberg et al, 1979) suggest 
that photosystem II of photosynthesis 
is heterogeneous with regard to its 
physical makeup and photochemical 
reactions, and it was suggested that 
the slow (0) component originates from 
the photosystem II complexes exposed 
in the stroma membranes whereas the 
fast (a) component originates from the 
photosystem II complexes in the ap- 
pressed membranes (Melis and Thielen, 
1980). The biphasic reoxidation kinet- 
ics of Q at 77 K shown in Fig. 24 may 
be a reflection of this physical hetero- 
geneity of the photochemical reaction 
centers of PS II, so that the fast and 
slow reoxidation components of Q oc- 
cur in the exposed stroma and the ap- 
pressed thylakoid membrane regions, 

The differences between these re- 
sults and those of Murata et al. (1973) 
may be caused by the use of different 
experimental material— leaves vs. ex- 
tracted chloroplasts. Leaves could be 
expected to have a more intact func- 
tioning photochemical system II and 
thylakoid aggregation compared to ex- 
tracted chloroplasts. 


Amesz, J., in Research in Photobiology, 
121-128, A. Castellani, ed.. Plenum Press, 
New York, 1977. 

Duysens, L. N. M., and H. E. Sweers, in Studies 
in Microalgae and Photosynthetic Bacteria, 
353-372, Japanese Society of Plant Physiolo- 
gists, ed., Tokyo, 1963. 

Melis, A., and L. N. M. Duysens, Photochem. 
PhotobioL, 29, 373-382, 1979. 

Melis, A., and P. H. Homann, Photochem. Pho- 
tobioL, 21, 431-437, 1975. 

Melis, A., and P. H. Homann, Photochem. Pho- 
tobioL, 23, 343-350, 1976. 

Melis, A., and A. P. G. M. Thielen, Biochim. Bio- 
phys. Acta, 589, 275-286, 1980. 

Murata, N., S. Itoh, and M. Okada, Biochim. 
Biophys. Acta, 325, 463-471, 1973. 



Rijgersberg, C. P., A. Metis, J. Amesz, and J. A. 
Swager, in Chlorophyll Organization and 
Energy Transfer in Photosysthesis, 305-322, 

Ciba Foundation Symposium 61 (New Series), 
Excerpta Medica, Amsterdam, Oxford, New 
York, 1979. 

GREEN ALGA Synechococcus lividus 

Gunnar Oquist, David C. Fork, Siegrid Schoch, and Gunilla Malmberg 

Heat inactivation of photosynthesis 
may be related to the breakage of short- 
range bonds of hydrophobic or hydro- 
philic character in photosystem II, as 
manifested in the functional disor- 
ganization of photosystem II at high 
temperatures (Schreiber and Armond, 
1978). It was of interest therefore to in- 
vestigate whether adaptation to high 
temperature in plants leads to an in- 
creased strength of the chlorophyll-pro- 
tein interactions, as reflected in the 
heat stability of the sodium dodecyl 
sulfate (SDS) solubilized chlorophyll- 
protein complexes of photosystem I 
and II, CP-a T and CP-a n . We used the 
thermophilic blue-green alga Synecho- 
coccus lividus because it contains only 
CP-a! and CP-an ( no light-harvesting 
CP-a/b complex, which overlaps CP-a n ) 
and also because considerable physio- 
logical data is already available on the 
photosynthetic properties of this alga 
(Fork et al, 1979). 

Synechococcus lividus (strain SY-4 
from Mercedes Edwards) was grown at 
38° and 55°C, as described previously 
(Fork et al, 1979). Thylakoids were iso- 
lated by a lysozyme treatment (modi- 
fied after the techniques reported by 
Ono and Murata [1979]), followed by 
French pressure cell treatment (77 kg 
cm -2 ), and harvesting of thylakoids by 
centrifugation at 48,000g- for 40 min. 
The thylakoids were solubilized for 5 
min in SDS at temperatures between 
0° and 65° C, and SDS-polyacrylamide 
gel electrophoresis (PAGE) and molecu- 
lar weight calibration were performed at 
4°C, as described earlier (Oquist and 

Samuelsson, 1980). The proportion of 
chlorophyll in the different bands of the 
gel was calculated by planimetry. Low- 
temperature absorption and fluores- 
cence emission spectra were measured 
and deconvoluted into components by 

Figure 25 shows gel scan profiles of 
extracts obtained at 20° and 55 °C from 
thylakoids of Synechococcus grown at 
55 °C. When grown at this tempera- 
ture, the alga has a lipid phase transi- 
tion temperature near 43 °C (Fork et al, 
1979). Similar profiles were obtained 
from algae that were grown at 38 °C 
(phase transition temperature near 
37 °C [Fork et al, 1979]). The profile ob- 
tained from thylakoids that were solu- 
bilized above the phase transition tem- 
perature had bands of CP-aj and CP-a n 
(apparent molecular weights of 85,000 
and 43,500, respectively), whereas ex- 
tracts made below the phase transition 
temperature showed only traces of CP- 
aj. Instead, a new band, CP-aj', ap- 
peared that had a slightly lower mobil- 
ity (apparent molecular weight 90,000- 
98,000) than CP-an 

Figure 26 shows the relative distribu- 
tion of CP-a r ', CP-an CP-a„ and SDS- 
Chl (solubilized chlorophyll) in the scan 
profiles obtained after SDS-PAGE of 
SDS-extracts made at temperatures be- 
tween 0° and 65 °C for Synechococcus 
grown at 38 °C (Fig. 26A) and at 55 °C 
(Fig. 26B). Clear bands of CP-a! were ob- 
tained from SDS-extracts made above 
40 °C, near the phase transition temper- 
ature of the 55°C-grown cells (Fig. 
26B). There was a tendency for the CP- 













2 4 6 8 

Distance from origin, cm 

Fig. 25. Densitometer tracings (677 nm) of 
chlorophyll-protein complexes isolated by SDS- 
PAGE from thylakoids of Synechococcus lividus 
grown at 55 °C and solubilized with SDS at 20 °C 
(lower curve) and 55 °C (upper curve) for 5 min. 

&i to appear at a few degrees lower 
temperature in 38°C-grown algae (Fig. 
26A). CP-a T of both algal cultures 
started to deteriorate when the solubil- 
izing temperature ranged above 60 °C. 
CP-aj', unlike CP-a T , was solubilized at 
temperatures between 0° and 40 °C. 
Above this temperature it started to 
disappear. Thus, CP-ai' disappeared 
from the scan profile as CP-a! ap- 
peared. Although under optimal condi- 
tions both CP-a r ' and CP^ made up 
40-50% of the chlorophyll on the gel, 
the chlorophyll content of CP-a T ' never 



38 °C Synechococcus 

o o CP-a, 

• • CP-a, 

4 4 CP-a,, 

■ SDS-Chl 

20 40 

Temperature, °C 

Fig. 26A. Relative distribution of chloro- 
phyll-protein complexes after SDS-PAGE of 
SDS extracts obtained by solubilizing thylakoids 
for 5 min at temperatures between and 65 °C. 
Synechococcus lividus was grown at 38°C. 




55 °C Synechococcus 

o o CP-a', 

• • CP-a, 

* » CP-a„ 


20 40 60 

Temperature, °C 

Fig. 26B. Same as Fig. 26A for Synechococ- 
cus lividus grown at 55°C. 



exceeded from one-third to one-half of 
the chlorophyll content in CP-aj (com- 
pare Fig. 25). The reason for this dis- 
crepancy between the presentations in 
Fig. 25 and Figs. 26 A and 26B, is that 
much less chlorophyll entered the gel if 
the extracts were obtained below 
rather than above the phase transition 
temperature. Extracts obtained below 
the transition temperature always gave 
rise to considerable amounts of chloro- 
phyll, which showed a tendency to 
enter and remain at the very top of the 
gel, whereas no (or only very little) resi- 
due was obtained with extracts solubil- 
ized above the phase transition temper- 
ature. When present, this residue had 
absorption and fluorescence emission 
properties typical for the antennae of 
photosystem I (data not shown). 

CP-a n , unlike CP-aj, was solubilized 
at high yield even at 0°C, and it started 
to decrease in the scan profile above 
about 40 °C. It could scarcely be found 
in extracts solubilized at 55-65 °C. The 
relative proportion of SDS-Chl stayed 
at 15-25% as long as the solubilization 
was performed below about 40 °C. With 
increasing solubilization temperature, 
it began an exponential rise as CP-an, 
and later CP-aj, started decreasing in 
the scan profiles. 

Low-temperature absorption and flu- 
orescence emission spectra were typ- 
ical for CP-a! and CP-a n , and no 
specific effects of the two growth tem- 
peratures on these properties were ob- 
served (data not shown). CP-a^ and CP- 
aj had identical absorption properties, 
and as they both had in addition a dom- 
inating far-red emission band we con- 

clude that they are both derived from 
the reaction center antennae of photo- 
system I. 

The phase transition temperature 
was considered to be critical for the sol- 
ubilization of CP-a I; either because of 
the difficulties of SDS (especially as it 
forms micelles at low temperatures) in 
penetrating the solidified membrane 
lipids at temperatures below the phase 
transition or because the CP-a T mono- 
mers of the photosystem I antennae 
are so strongly bound to each other 
that they cannot be dissociated by 
SDS before thermal agitation has 
reached a certain level, achieved only 
above the phase transition tempera- 
ture. Below this temperature, only a 
fraction of the photosystem I antennae 
(CP-ai') is accessible to SDS. Although 
CP-a n of Synechococcus could be solu- 
bilized at high yield even at 0°C, it was 
remarkably heat-stable in comparison 
with what is normal for mesophilic 
plants. We consider the difficulties in 
solubilizing CP-a! at sub-transition 
temperatures, and the heat stability of 
the two complexes, both as adapta- 
tions enabling Synechococcus to grow 
under extreme high temperature re- 


Fork, D. C, N. Murata, and N. Sato, Plant 

Physiol., 63, 524-530, 1979. 
Ono, T., and N. Murata, Biochim. Biophys. 

Acta, 545, 69-96, 1979. 
Oquist, G., and G. Samuelsson, Physiol. 

Plant, 50, 57-62, 1980. 
Schreiber, U., and P. A. Armond, Biochim. 

Biophys. Acta, 502, 138-151, 1978. 






Kazuhiko Satoh and David C. Fork 

It is well known that under aerobic 
conditions chlorophyll fluorescence 
shows a typical induction phenomenon 
in leaves, algal cells, or intact chloro- 
plasts (Kautsky and Appel, 1960; 
Govindjee and Papageorgiou, 1971; 
Katoh et al, 1975). However, under 
anaerobic conditions this typical in- 
duction pattern of chlorophyll fluores- 
cence is replaced by another, simpler 
one. For example, with short incuba- 
tion of photosynthetic organelles undey 
anaerobic conditions, the PS X decline is 
largely inhibited, and sometimes a slow 
fluorescence decay after the peak P is 
observed (Bannister and Rice, 1968; 
Schreiber and Vidaver, 1974). It is also 
known that anaerobic treatment of 
algal cells inhibits the first phase of the 
induction of oxygen evolution (Ban- 
nister and Rice, 1968; Schreiber and 
Vidaver, 1974). 

Almost all work concerning the ef- 
fects of anaerobiosis on the fluores- 
cence time course has been done using 
complex systems such as leaves or 
algal cells. Moreover, little work on the 
slow fluorescence decline which replaces 
the PSiMj transient has been done un- 
til now. Since oxygen was found to act 
as the main electron acceptor after 
photoactivation of electron transport 
had taken place (Satoh, this Report), it 
was also interesting to investigate 
what would happen if chloroplasts were 
illuminated in the absence of oxygen. 

In this study, we used intact Bryop- 
sis chloroplasts and observed the ef- 
fects of anaerobic treatment on the 
time course of chlorophyll fluorescence 
and on the electron transfer reactions. 
We found that the fluorescence decline 
was related to photoinhibition of the 
electron transport system. 

Figure 27 shows effects of anaero- 
biosis on the time course of chlorophyll 

fluorescence at low and high light in- 
tensities. In order to obtain anaerobic 
conditions, 5 mM dithionite (Fig. 27, 
b and e) or glucose + glucose oxidase 
(Fig. 27, c and f) were added. Under 
aerobic conditions, the typical DPSiMj 
transient of chlorophyll fluorescence 
was observed (Fig. 27, a). The fluores- 
cence decline from the peak M x was 
very slow even at the higher light inten- 
sities used (Fig. 27, d). However, under 
anaerobic conditions, this PSjMx tran- 
sient was replaced by a simpler decline 
from the peak P, and fluorescence 
reached much lower levels than those 
obtained under aerobic conditions. 
Under iso-osmotic conditions (corre- 
sponding to 1.0 M sorbitol), electron 
carriers were not reduced by the addi- 
tion of dithionite in intact Bryopsis 
chloroplasts (data not shown). There- 
fore, the effects of dithionite addition 
can be attributed to the consumption of 
oxygen by dithionite in the reaction 
mixture rather than to its effects on 
electron carriers directly. This fluores- 
cence decline was also observed in chlo- 
roplasts incubated with glucose and 
glucose oxidase, although the rates 
were lower in this case (Fig. 27, c and f). 
These results suggest that under an- 
aerobic conditions the electron trans- 
port pathway of the chloroplasts was 
modified so that a simpler fluorescence 
decline replaced the PSiMj transient. 

Under conditions of low C0 2 concen- 
tration, light is known to induce inhibi- 
tion of photosynthesis (Powles and Os- 
mond, 1979). This photoinhibition is 
attributed to the presence of excess 
light energy because there may be es- 
sentially no acceptor pools for elec- 
trons under low C0 2 conditions. As 
mentioned previously, oxygen was 
shown to act as a principal electron ac- 
ceptor after photoactivation of electron 




CD ^ 










3 ff on 






Fig. 27. Time courses of chlorophyll fluorescence under aerobic and anaerobic conditions in intact 
Bryopsis chloroplasts: curves a and d, under aerobic conditions; b and e, +5 mM dithionite; c and f, 
+ 10 mM /3-D-glucose, 20 units/ml of glucose oxidase and 60 units/ml of catalase. Upper line (a, b, and c), 
at a low fluence rate of light (2.0 X 10 4 erg cm -2 s~ l ). Lower line (d, e, and f), at a higher fluence rate of 
light (2.0 X 10 5 ergcm" 2 s _1 ). Reaction mixture contained, in 2 ml, 1.0 M sorbitol, 50 mM HEPESfpH 
7.5), 11 mM MgCl 2 , 1 mM MnCl 2 , 2 mM EDTA, 2 mM NaN0 3 and chloroplasts equivalent to 5.15 ^g 

transport. If the above hypothesis is 
true, then in the absence of oxygen al- 
most any light intensity may bring 
about photoinhibition. Therefore, we 
measured activities of photosystem I 
and II after illumination of the chloro- 
plasts under anaerobic conditions. The 
activity of photosystem I (methyl vi- 
ologen photoreduction in the presence 
of DCMU and DCIPH 2 ) decreased 
slowly and was about 87% of its initial 
activity after 45 s of illumination (Fig. 
28, square symbols). On the other 
hand, the activity of photosystem II 
(DCIP-Hill reaction, Fig. 28, circles) de- 
creased sharply and then continued to 
decrease slowly. The fluorescence time 
course under anaerobic conditions is 
also shown in Fig. 28. It also has two 
phases, a rapid and a slow phase. The 
time course of the initial rapid phase of 

the activity decrease of photosystem 
II was similar to that of the rapid 
phase of the fluorescence decline. Illu- 
mination of the chloroplasts under aer- 
obic condtions had no effect on the pho- 
tosystem I and II activities (data not 
shown). These results suggest that the 
rapid phase is directly related to photo- 
inhibition of photosystem II, and 
although the rates are different, the 
slow phase is concerned with photoin- 
hibition of both photosystems. 

If we accept the hypothesis that oxy- 
gen is a main electron acceptor after 
the photoactivation has taken place, 
this photoinhibition must be induced 
by low intensities of light under anaer- 
obic conditions. Table 6 shows the ef- 
fect of various intensities of actinic 
light on the rate of the DCIP-Hill reac- 
tion. The illumination period was also 



1 1 1 1 I 
V "" ~ ~°" ~ - ~ 

\ ~ ~ ~ 



••■... A 

— A ■ ■ ■ 


1 1 1 1 1 


TABLE 6. Effects of Various Intensities of 
Preillumination Light on the Rate of the DC IP- 
Hill Reaction.* 


50 - 

25 - 

10 20 30 40 50 60 

Preillumination time, sec 

Fig. 28. Effects of preillumination periods 
under anaerobic conditions on the photosystem I 
and II reactions. Photosystem I reaction 
(squares) was measured by oxygen consumption 
by methyl viologen photoreduction in the pres- 
ence of DCMU and DCIPH 2 (as an electron 
donor). Concentrations of methyl viologen, 
DCIPH 2 , DCMU, and chlorophyll were 2 mM, 0.2 
mM, 10 fiM, and 50 uglml, respectively. The in- 
tensity of the actinic light was 1.7 X 10 4 lux. 
Photosystem II reaction (circles) was measured 
by the light-induced absorbance decrease of 
DCIP with water as an electron donor. Concen- 
trations of DCIP, methylamine, and chlorophyll 
were 0.2 mM, 30 mM, and 10 ^g/ml, respectively. 
The intensity of the actinic light was 2.8 X 10 5 
erg cm' 2 s^ 1 . The fluorescence time course 
(triangles) was measured during the preillumina- 
tion of the chloroplasts. Other conditions were 
the same as in Fig. 27b. 

changed to observe the effect under 
conditions in which the same numbers 
of photons were absorbed by the 
chloroplasts. Light even at intensities 


Light Intensity 


(erg cm -2 s _1 ) 

Time (s) 

Reaction (%) 


7.7 X 10 4 



7.7 X 10 3 



7.7 X 10 2 



* Chloroplasts were preilluminated in the 
presence of 5 mM dithionite and then washed 
twice to remove the dithionite. Intensities of the 
preillumination light and the periods of preil- 
lumination were as indicated in the Table. 0.2 
mM DCIP and 30 mM methylamine were added 
to the reaction mixture. Chlorophyll concentra- 
tion was 16 jug/ml. Other conditions were the 
same as in Fig. 28. 

as low as 7.7 X 10 2 erg cm -2 s -1 induced 
photoinhibition of photosystem II. 
These results also show that this sys- 
tem may provide a good model to 
study the initial event of photoinhibi- 
tion of photosynthesis. 


Bannister, T. T., and G. Rice, Biochim. Biophys. 
Acta, 162, 555-580, 1968. 

Govindjee, and G. Papageorgiou, in Photo- 
physiology, Vol. 6, 1-46, A. C. Giese, ed., 
Academic Press, New York, 1971. 

Katoh, S., K. Satoh, A. Yamagishi, and T. Ya- 
maoka, Plant Cell Physiol, 16, 1093-1099, 

Kautsky, H., and W. Appel, Biochim. Z., 332, 
277-292, 1960. 

Powles, S. B., and C. B. Osmond, Plant Physiol, 
64, 982-988, 1979. 

Schreiber, U., and W. Vidaver, Biochim. Bio- 
phys. Acta, 368, 97-112, 1974. 



David C. Fork, Gunnar Oquist, and Stephen B. Powles 

Photoinhibition of photosynthesis re- 
sults from exposure of plants to exces- 
sive light energy. The first signs of 
such damage are a reduction of the 
quantum yield for photosynthetic C0 2 

uptake accompanied by inhibition of 
PS II electron transport and loss of flu- 
orescence emission from PS II (Powles 
et al, 1979; Powles and Critchley, 
1980). These results suggest that pho- 



toinhibition results as a consequence 
of over-energization of the photosyn- 
thetic reaction centers. In other stud- 
ies, we have been examining how desic- 
cation alone, and in combination with 
light, affects energy distribution in 
photosynthesis in marine algae, liver- 
worts, and lichens (Oquist and Fork, 
Fork and Oquist, this Report; Year 
Book 79, 193-197). Even in these stud- 
ies, a preferential loss of fluorescence 
emission from PS II was induced. In 
measurements with algae, liverworts, 
and lichens, the preferential loss of flu- 
orescence emission from PS II could be 
explained by (1) changes in a £77//-/) 
(spillover of light energy from PS II to 
PS I), (2) changes in a (the fraction of 
light distributed originally to PS I), (3) 
changes in kF n (the rate constant for 
PS II fluorescence), or (4) by reaction 
center quenching produced by accumu- 
lation of inactive reaction centers 
(P680+ Q-) of PS II (see Butler, 1978). 
In this study we wanted to determine 
whether the preferential loss of PS II 
emission under photoinhibition treat- 
ments could be explained by any of the 
above factors and if so by which one(s). 

Bean plants (Phaseolus vulgaris C v 
Hawkesbury Wonder) were grown un- 
der 25°/20°C temperatures during the 
late summer and autumn months when 
light intensities regularly reached quan- 
tum fluxes of 2000 fimol quanta m~ 2 
s _1 . Photoinhibition was accomplished 
by treating intact attached fully ex- 
panded bean leaflets for 3 h at 6°C 
under white light at 2000 jumol quanta 
m- 2 s" 1 (see Year Book 79, 157-160, for 
description of treatment procedures). 

Measurements of fluorescence emis- 
sion spectra were done at 77 K, as de- 
scribed prevously (this Report). Fluo- 
rescence was measured simultaneously 
at 695 and 740 nm using a trifur- 
cated fiber optic light guide. One 
branch served to illuminate the sample. 
The other two branches were fitted 
with interference filters (Infra Red 
Industries, 5-nm half-bandwidth) and 
attached to photomultipliers (EMI 

9558B and Hamamatsu TV R928). The 
amplified signals were each fed to the 
inputs of a two-pen strip chart re- 

A quartz light guide was attached to 
the optic fiber bundles and served to 
hold the sample submerged in a Dewar 
of liquid nitrogen while guiding actinic 
light to the leaf surface and collecting 
fluorescence from the same surface. 
The intensity of the actinic light was 
decreased until it became possible to 
follow the kinetics of fluorescence at 
77 K. 

Figure 29, curve A, shows the emis- 
sion spectrum of a bean leaflet that had 
been kept at 6 °C for 3 h in the dark (in 
normal air). In Fig. 29, curve B, the in- 
tact bean leaflet was given a 3-h ex- 
posure to 6°C (normal air) at a photon 
fluence rate of 2000 /miol quanta m~ 2 

4 - 

-i — i — i — I — i — i — i — i — I — i — i — i — i — I — i — i — i — r 





700 750 800 

Wavelength, nm. 


Fig. 29. Low-temperature (77 K) fluorescence 
emission spectra for an intact bean leaflet that 
had been kept in the dark for 3 h at 6°C (A) and for 
a leaf kept at 6°C in white light having a photon 
fluence rate of 2000 /umol quanta m ~ 2 s ~~ ' (B). The 
fluorescence was sensitized by chlorophyll b ex- 
citation (5 /xmol quanta m~ 2 s -1 , obtained with a 
Balzers interference filter 478 nm, half- 
bandwidth 9 nm, plus Corning 5-60 and Calflex C 
filters. The slits of the monochromator were set 
to pass a half-bandwidth of 1.6 nm. 



s _1 (full sunlight). These spectra show 
that exposure of a bean leaflet to a high 
light intensity while at a chilling tem- 
perature results in a preferential de- 
crease of the PS II fluorescence peaks 
at 687 and 695 nm. The ratio of fluores- 
cence peak height at 740(PS I) to that 
at 695 nm was 3.6 for the control and 
5.4 for the photoinhibited leaflet. The 
ratios of 740/685 fluorescence were 4.2 
for the control and 7.0 for the pho- 
toinhibited leaflet. 

The preferential loss of PS II activity 
in the photoinhibited leaflet was not 
seen when the fluorescence emission 
spectrum of the underside of the same 
leaflet was measured (data not shown). 
Since the upper surface of the treated 
leaflet was maintained perpendicular to 
the light during the photoinhibition 
treatment, the chloroplasts in cells to- 
ward the lower leaflet surface were 
shielded from the light energy imping- 
ing upon the plastids in the upper 
layers of the leaflet. There is, therefore, 
probably a gradient of damage to the 
plastids through the leaflet. One conse- 
quence of this gradient, as also pointed 
out by Critchley (1981), is that prepara- 
tions of chloroplasts isolated from leaf 
material photoinhibited in this way will 
contain plastids from throughout the 
leaf. Therefore, damage exhibited by 
chloroplasts close to the upper exposed 
leaf surface will be underestimated. 

This is exemplified in Fig. 30, which 
shows emission spectra for chloro- 
plasts prepared from a bean leaflet pho- 
toinhibited at 6°C, as described above. 
The photoinhibition effect is much less 
pronounced in the chloroplast prepara- 
tion than observed from the exposed 
upper leaf surface (Fig. 29, curve B). 
The ratios of the 741/697 peak heights 
are 2.3 for the control and 2.5 for the 
photoinhibited. The 741/688 ratios are 
2.2 for the control and 2.4 for the photo- 
inhibited chloroplasts. 

Kinetics of the rise of fluorescence at 
695 and 740 nm were measured in a 
control leaflet in a photoinhibited (at 
6°C) leaflet at 77 K in order to deter- 

«= 4 

= 2 

"i i r~| i i i i | — i — i — i — i — I — i — i — i — r 





700 750 800 

Wavelength, nm. 


Fig. 30. Low-temperature (77 K) fluorescence 
emission spectra for chloroplasts isolated from 
control and photoinhibited bean leaflets using 
chlorophyll b excitation, as described in Fig. 29. 

mine if the loss of PS II emission can be 
accounted for by one or more of the 
four possibilities mentioned above. 

It has been demonstrated that Fj = 
f(F n ) when the PS II traps are all 
closed (reduced) at 77 K (Kitajima and 
Butler, 1975). This function gives a 
straight line in an X-Y plot. It can be 
shown that the slope of the line 

Fi/F u = k-nn^n'tFjIkFu. 

The intercept obtained when this 
straight line is extrapolated back to the 
Y-axis is proportional to a, the fraction 
of light energy absorbed by PS I di- 
rectly and independently of the state of 
the PS II traps (Butler, 1978). 

Figure 31 gives the plots of PS I (at 
740 nm) and PS II emission (at 695 nm) 
during trap closure of a control and of a 
photoinhibited bean leaflet. Figure 31 
shows results obtained by measuring 
the upper surface and the under surface 
of the photoinhibited leaflet. The upper 
side of the photoinhibited leaflet had a 
strong decrease of F v particularly at 
695 nm since F m decreased from about 
150 to 60, but F of the photoinhibited 
leaflet did not change in comparison to 
the control leaf. Photoinhibition, in ad- 
dition, produced about a 45% increase 



Fig. 31. F695 vs. Fl 40 measured at 77 K for 
the upper side of control and photoinhibited bean 
leaflets (A) and for the under side of control and 
photoinhibited leaflets (B). Lowest point for each 
curve represents F \ highest point, F m \ 
F , — F - F . 

in the slope of the line and an a decrease 
from 120 to 113 units. 

The observation that the PS II fluo- 
rescence was quenched specifically in 
F v and not in F indicates that chloro- 
plasts in the bean leaflets exposed to 
photoinhibitory treatment exhibited 
reaction center quenching produced by 
accumulation of P680 + Q~. It is charac- 
teristic of this type of quenching that it 
affects only F v but not F , a conclusion 
indicated by studies of UV effects 
(Malkin and Jones, 1968; Yamashita 
and Butler, 1968) and ferricyanide ef- 
fects on photosynthesis (Okayama and 
Butler, 1972; Butler, 1978). The partial 
loss of F v 740 can also be explained by 
P680+Q - quenching, which would de- 
crease the amount of excitation energy 
spilled over to PS I. It has been con- 
vincingly shown before that the var- 
iable portion of PS I fluorescence at 
77 K is a function of the state of the 
traps of PS II (Butler, 1978). 

The increased slope of the line pro- 
duced by photoinhibition (curve A, Fig. 
31) may be a sign of increased spillover, 
^T(//-/). but it could also be caused by 
an increase in the quantum yield of 
PS I fluorescence (4>Fj), which was seen 
to increase by about 15% in a test ex- 
periment (excitation at 700 nm). The 
function k,F H probably does not change, 
since F was unaffected. However, 
more studies will be needed to deter- 
mine if the slope change induced by 
photoinhibition is produced by in- 
creases in k^jj^j) or 4>F I or both. It can 
be seen in Fig. 31, curve B, that the 
under surface of the leaflet responded 
much less than did the upper surface to 
the photoinhibitory treatment. These 
results confirm the emission-spectra 
findings and demonstrate clearly that 
chloroplasts in cells of the upper sur- 
face of the leaflet became preferentially 

Effects similar to those described for 
Fig. 31, curve A, were seen in chloro- 
plasts isolated from photoinhibited 
leaflets except that they were much 
less pronounced. This result, like the 
result obtained for emission spectra 
shown in Fig. 30, would be caused by 
the chloroplast preparation containing 
a mixture of photoinhibited and non- 
photoinhibited thylakoids. 

Measurements of Fj and F H fluores- 
cence at 77 K from water-stressed and 
sun-exposed or water-stressed and 
shaded Nerium oleander leaves also in- 
duced a specific quenching of F v . This 
reaction center quenching was stronger 
in water-stressed, light-exposed leaves 
than in water-stressed, shaded leaves 
(this Report), and it was correlated 
directly to the loss of the capacity for 
electron transport in PS II {Year Book 
79, 150-157). 

It would appear from these and other 
studies that photoinhibition of photo- 
synthesis can occur after photosynthe- 
sis is inhibited either directly (e.g., by 
removal of C0 2 ) or by environmental 
stress (e.g., temperature and/or water 
stress). As demonstrated above, photo- 



inhibition effects can be observed from 
X-Y plots of PS I vs. PS II fluores- 
cence, as had been done in the desicca- 
tion studies of algae (Oquist and Fork, 
1981). In order to test further the 
hypothesis that photoinhibition can oc- 
cur as soon as photosynthesis is in- 
hibited, we measured X-Y plots of PS I 
vs. PS II fluorescence in leaf discs that 
were infiltrated with DCMU and 
floated on water in the presence or 
absence of light. Figure 32, curve A, 
shows that the Fj vs. F H plots of dark 
control leaf discs ± DCMU-infiltration 
were identical; both lines had the same 
slopes, the same F m values, and about 
the same F values. By contrast, a 
5-min exposure of the DCMU-infil- 
trated leaflet to white light produced a 
marked quenching of F v (Fig. 32, curve 
B) analogous to the effects of the high 
light, chilling-induced effects on F v in 
bean leaflets (compare Fig. 31, curve 
A). However, photoinhibition in the 
presence of DCMU produced no slope 
or a changes but caused an increased 
F . The increase in F may have been 
produced by a fraction of the reaction 
centers that did not become reoxidized 
during the 1-5 min dark periods which 
followed the photoinhibitory treatment 
before freezing to 77 K. 

The preferential loss of PS II emis- 
sion observed in photoinhibition stud- 
ies at room temperature (Critchley and 
Smillie, 1981), at 77 K (Powles et al, 
1979), and in this study can be ascribed 
to some type of damage to PS II. Pho- 
toinhibition effects could not be allevi- 
ated by adding to PS II electron do- 
nors that can bypass the water-split- 
ting complex (Critchley, 1981). This 
result suggests that the site of damage 
lies at or very close to the PS II re- 
action center. The present results sug- 
gest that the primary damage done by 
photoinhibition must be in the immedi- 
ate environment of P680 or in the pro- 
tein, carotenoids, or neighboring lipids 
with which P680 is associated (see 
Satoh and Fork, this Report). 

The loss of F v in photoinhibited 



J I 1 i i 

-I I 1 L_ 

50 100 150 20( 

Fluorescence 695, rel 

Fig. 32. F695 vs. F740 measured at 77 K for 
dark control and DCMU-infiltrated dark-adapted 
bean leaf discs (A). In (B), the DCMU-infiltrated 
leaf discs were photoinhibited by a 5-min ex- 
posure to white light (2000 ^mol quanta m~ 2 
s -1 )- See Fig. 31 legend for details. 

leaves may be caused by the creation of 
a more oxidizing environment that 
allows accumulation of oxidized P680. 
Such a condition is conducive to an 
interaction between excitation energy 
and oxygen, resulting in the formation 
of singlet oxygen. This species of ox- 
ygen is thought to be very active in hy- 
drophobic environments such as PS II 
(Fridovich, 1976), where it can initiate 
potentially destructive oxidative reac- 


Butler, W. L., Annu. Rev. Plant. Physiol., 29, 

345-378, 1978. 
Critchley, C, in Proc. Fifth Int. Photosynthesis 

Congress, in press, 1981. 



Critchley, C, and R. M. Smillie, Aust. J. Plant 

Physiol., in press, 1981. 
Fridovich, I., in Free Radicals in Biology, Vol. 1, 

239-277, W. A. Pryor, ed., Academic Press, 

New York, 1976. 
Kitajima, M., and W. L. Butler, Biochim. 

Biophys. Acta, 376, 105-115, 1975. 
Malkin S., and L. W. Jones, Biochim. Biophys. 

Acta, 162, 297-299, 1968. 
Okayama, S., and W. L. Butler, Biochim. 

Biophys. Acta, 267, 523-529, 1972. 

Oquist, G., and D. C. Fork, Plant Cell Physiol., 

in press, 1981. 
Powles, S. B., and Critchley, C, Plant Physiol., 

65, 1181-1187, 1980. 
Powles, S. B., C. B. Osmond, and S. W. Thorne, 

Plant Physiol., 64, 982-988, 1979. 
Yamashita, T., and W. L. Butler, Plant Physiol, 

43, 2037-2040, 1968. 



IN Nerium oleander 

Olle Bjorkman, Stephen B. Powles, David C. Fork, and Gunnar Oquist 

Last year we reported on the effect of 
water stress on the photosynthetic 
characteristics of Nerium oleander 
{Year Book 79, 150-157). It was found 
that next to stomatal responses, photo- 
synthetic electron transport and photo- 
phosphorylation are the component 
processes most sensitive to water 

Our preliminary observations in- 
dicated that high irradiance levels may 
aggravate the detrimental effects of 
water stress. During the past year we 
have therefore studied the interactions 
between irradiance level and leaf water 
status on photosynthetic reactions in 
N. oleander plants. 

For this purpose, plants were grown 
in soil (18-1 containers) in the Depart- 
ment's experimental garden during 
spring and summer of 1980. In late 
summer, water stress was imposed on 
some of these well-established plants 
by withholding the water supply to the 
soil. The leaf water potential (\}/J, 
monitored several times weekly, was al- 
lowed to fall from about —1.0 to —6.0 
MPa over a one-month period. Plants 
serving as controls were supplied with 
ample water throughout the experi- 
ment. One set of stressed plants and 
control plants received full daylight ir- 
radiance levels while another set was 
placed in the shade, receiving 5-10% of 
full daily irradiance. All leaves used for 

experiments were kept horizontal 
throughout the treatment period. Leaf 
water potential, stomatal conductance, 
chlorophyll content, coupled and un- 
coupled electron transport, and the 
fluorescence kinetics at liquid nitrogen 
temperature were followed periodically 
on stressed and control plants during 
the treatment period. The methods used 
here have been previously described 
(Year Book 79, 150-157; Fork, Oquist, 
and Powles, this Report). 

Some of the results of these studies 
are summarized in Tables 7 and 8. The 
values for whole-chain electron trans- 
port by isolated chloroplasts and 
variable fluorescence (F v ) at 695 nm 
from stressed leaves are expressed as 
percentages of those obtained from 
concurrent measurements on control 

It is evident from Table 7, that in 
plants exposed to the full daylight 
levels under which the leaves had de- 
veloped, the rate of uncoupled whole- 
chain electron transport showed a pro- 
gressive decline as the leaf water 
potential decreased. The control val- 
ues, as well as the chlorophyll content 
of both control and stressed plants, did 
not show any marked variation during 
the treatment period. Photosystem 
I -driven electron transport was much 
less affected by low leaf water poten- 
tial than whole-chain electron trans- 



TABLE 7. Effect of Leaf Water Potential (i£J 
on the Uncoupled Rate of Whole-Chain Electron 
Transport (ET) of Isolated Chloroplast Mem- 
branes of Nerium oleander, Maintained in the 
Sun and in the Shade 


, % of Watered Control* 

*„. MPa 


Photosystem I 





















*For watered control plants, i/^ ranged from 
—0.5 to —1.2 MPa and the rate of electron 
transport (H 2 O — MV) for these plants was 107 
±6 (sun) and 79 ± 17 (shade) ^mol 2 g -1 Chi 
s _1 at25°C. 

TABLE 8. Effect of Leaf Water Potential tyj 
on Variable Fluorescence (F v ) from the Upper 
Surface of Leaves of Nerium oleander, Main- 
tained in the Sun and in the Shade 

F v , % of Watered 

>P W , MPa 
















* Fluorescence emission at 695 nm was 
measured at liquid nitrogen temperature, as 
described by Fork, Oquist, and Powles (this 
Report). For the present purpose we define 
variable fluorescence as F v = (F m — FJ/F , 
where F m and F denote the maximum and the 
initial fluorescence yields, respectively. The F v 
values for the control plants were 2.4 ± 0.2 (sun) 
and 4.7 ± 0.4 (shade). It should be noted that 
because of light gradients in the leaf, high-light- 
induced inactivation is likely to have a greater 
effect on the fluorescence characteristics of the 
upper leaf surface than on the rate of electron 
transport of a population of chloroplasts iso- 
lated from exposed leaves. 

port. Also, there was no marked 
change in the ratio between the rates in 
the presence and in the absence of added 
uncoupler, either for whole-chain or 
photosystem I -driven electron trans- 
port (data not shown). As shown in the 
lower part of Table 7, the effect of \j/ u on 
whole-chain electron transport was 
much smaller when the leaves were 
kept in the shade than when they were 
exposed to full daylight. Much lower 
water potentials were required to pro- 
duce a significant inhibition of electron 
transport, and at any given water po- 
tential this inhibition was much less 
pronounced in the shade than in full 

As shown in Table 8, plants exposed 
to full daylight at low leaf water po- 
tential exhibited a progressive decline 
in the variable fluorescence (F v ) emit- 
ted from photosystem II of leaves at 
695 nm. This decline in F v was primar- 
ily caused by a decrease in the F m level; 
there was no pronounced or consistent 
change in the F levels with changes in 
\J/ U . The fact that water-stressed leaves 
exposed to high irradiance had a speci- 
fic quenching of F v of photosystem II 
and not F indicates that this treat- 
ment led to an inactivation of photo- 
system II reaction centers as a result 
of a type of reaction center quenching 
that is discussed in another section of 
this Report (Fork, Oquist, and Powles). 

As shown in the lower part of Table 
8, water stress had only a small effect 
on F v over a very wide range of ^ w 
when the leaves were kept in the shade 
during the stress treatment period. 
These results are in close agreement 
with those obtained in the electron 
transport measurements on isolated 

The results of this study demon- 
strate that there is a close interaction 
between irradiance level and leaf water 
status on inactivation of photosyn- 
thetic reactions. The inactivation of 
photosystem II activity and the quench- 
ing of the photosystem II reaction 
centers were much more pronounced 



when water stress was combined with a 
high irradiance level than when it was 
combined with a low irradiance level. 
However, studies now in progress show 
that leaves of N. oleander grown under 
low or moderate irradiance levels suffer 
substantial photoinhibition when ex- 
posed to full daylight irradiance levels, 
even when the leaf water potentials re- 
main as high as —0.5 MPa. These pho- 
toinhibitory effects very closely resem- 
ble those resulting when field-grown N. 

oleander plants are subjected to water 

These results indicate that water 
stress, by some yet unknown mecha- 
nism, increases the susceptibility to 
photoinhibition of N. oleander leaves. 
One may speculate that this is the main 
agent by which water stress causes 
damage to the photosynthetic appa- 
ratus of plants under natural condi- 




Olle Bjorkman and Stephen B. Powles 

Oxalis oregana Nutt. (redwood sor- 
rel) is an ubiquitous and dominant com- 
ponent of the evergreen vegetation of 
the densely shaded floor of the red- 
wood {Sequoia sempervirens) forests of 
northern California. Although this Ox- 
alis species predominates in the most 
densely shaded sites occupied by 
higher plants (about 0.5% of full sun- 
light), it extends to the borders of 
forest clearings. In more open sites, the 
plants may be subjected to intense sun- 
flecks; in sites adjacent to clearings the 
quantum fluence rate may reach that of 
full sunlight and last for periods up to 
one or two hours. Under these condi- 
tions, changes in irradiance of about 
200-fold may occur within a few sec- 
onds, thus imposing a drastic shift in 
the radiation environment of the leaf. 

Like many other species, Oxalis 
oregana is capable of changing the 
orientation of its leaves, enabling it to 
maximize light interception as the 
direction of the light changes. Such 
maximization is advantageous in light- 
limited habitats. In Oxalis oregana and 
many other members of the Oxalida- 
ceae, the leaf consists of three leaflets, 
each of which is attached to a single 
vertical petiole by its own pulvinus. 
These organs, serving as swivel joints, 

permit each leaflet to change its angle 
(to the horizontal plane) over a range of 

Our field observations show that for 
most of the day, the three leaflets of 
Oxalis oregana are oriented in essen- 
tially the same plane, facing the bright- 
est part of the forest canopy (Fig. 33, 
left). However, when the leaves are 
struck by direct solar radiation, they 
respond by folding downward toward a 
position in which the leaflets are 
parallel to the sun's rays, thus minimiz- 
ing light interception (Fig. 33, right; 
this is also the nighttime position of the 
leaflets). This folding response is 
remarkably rapid. As shown in Fig. 34, 
the change in leaf angle, following a 
change in incident quantum fluence 
rate from approximately 4 to 1600 
/xmol quanta m~ 2 s _1 , is 20° per min. 
The lag is only about 10 s and the ap- 
proximately 90° change in leaf angle 
(equivalent to a change in light inter- 
ception from over 90% to less than 
10%) is complete in approximately 6 
min (Fig. 34; compare Fig. 33, right). 
The return of the leaves to an essen- 
tially horizontal position after depar- 
ture of the direct sunfleck is a con- 
siderably slower process. In the example 
given in Fig. 34, there was a lag of 



Fig. 33. Photographs showing the position of Oxalis oregana leaves immediately before (left) and 
6 min after (right) the plant was struck by an intense sunfleck. San Mateo Memorial Redwood Park, 
San Mateo County, California, April 13, 1981 (a clear day), Experimental Site #2. 

about 10 min, followed by a change in 
leaf angle at a rate of about 3 ° per min. 
The total time required for the leaves to 
return to the original position after 
departure of the sunfleck was 30-35 

The effect of such changes in leaf 
angle on the radiation environment of 
the leaves, photosynthetic perform- 
ance, and the role of leaf folding in 
minimizing light-induced injury to the 
photosynthetic system are considered 
in the following article. 

Subsequent studies both in the 
native redwood sites and in the labora- 
tory were undertaken to determine the 
relationship between quantum fluence 
rate and leaf folding, the action spec- 
trum for the folding response, and the 
sensory mechanism involved. A brief 
summary of some of the results are 
given below. 

The threshold light level for trigger- 
ing the rapid leaf-folding response in 
young leaves is the range of 300-400 
pimol quanta m~ 2 s _1 . Similar results 
were obtained with natural sunlight, 
and with xenon arc and Metal arc 
lamps as the light sources. Older leaves 
required somewhat higher light levels 
for triggering the folding reaction and 

they tended to respond more slowly 
than young leaves. Laboratory mea- 
surements on Oxalis plants, grown 
under natural shade in the garden at 
Stanford, showed that a minimum 
quantum fluence rate of 600 /^mol 
quanta m~ 2 s _1 was required to trigger 
the rapid folding reaction in 50% of the 
leaves of varying age for a population 
of 20 plants. This value may be 
somewhat higher than that for plants 
growing in their native habitat, per- 
haps because the average quantum 
fluency rate under which the leaves 
developed was several times higher in 
the shade garden than in a typical 
native site. 

Increases in the quantum fluence 
rate above the threshold level de- 
creases the lag time, increases the rate 
of leaf folding, and also increases the 
final change in leaf angle. For example, 
at a quantum fluence rate of 800 jumol 
m~ 2 s _1 , the lag time was 1 min, the 
rate of change in leaf angle was 10° per 
min, and the final change in leaf angle 
was 69° for a given leaf. Corresponding 
values for the same leaf at a quantum 
fluence rate of 1500 /xmol m -2 s _1 were 
20 s, 19° per min, and 89°. 

The primary sensory organ involved 













10 - 


70 - 

CD ! 

C 11 



1 1 1 1 1 1 1 1 1 1 1 1 I 1 1 1 1 


1* / 

\ /* 

1 1 1 1 1 •! 1 1 1 1 1 1 1 1 1 1 1 

12 24 36 

Time, minutes 


Fig. 34. Time course of leaf folding in Oxalis oregana in response to an intense sunfleck (1590 
/amol quanta m -2 s~\ horizontal plane) and of leaf unfolding after departure of the sunfleck. The 
quantum fluence rate in the shade was 4 /xmol quanta m~ 2 s _1 . Measurements were made on April 
14, 1981 (a clear day) on the plant stand shown in Fig. 33. 

in the rapid folding reaction of Oxalis 
oregana leaves is the pulvinus itself. 
No folding response was obtained 
when the pulvinus was shaded, even 
when the remainder of the leaf was il- 
luminated with bright light (2000 /xmol 
quanta m~ 2 s _1 ). Conversely, full 
response was obtained when the 
pulvinus alone was illuminated and the 
remainder of the leaf was shaded. Since 
illumination of one pulvinus while 
shading the other two pulvini elicited a 
response only in the leaflet attached to 
the illuminated pulvinus, it appears 
that the responses of the three leaflets 
are largely independent of one another. 
Although a leaf-folding reaction can 
be induced by mechanical agitation of 
the leaf or by rapidly changing the 
water relations of the leaf (e.g., by exci- 
sion of the petiole), the primary agent 
in eliciting the folding reaction is light, 
more specifically, blue light. When 
wavelengths shorter than 500 nm are 
removed from sunlight (by applying 
appropriate filters), the folding reac- 

tion is totally abolished, even when the 
total quantum fluence rate in the visi- 
ble region is kept as high as 2200 /xmol 
m -2 s _1 . Irradiation with wavelengths 
in the waveband 700-2400 nm also did 
not trigger a folding reaction even at 
intensities considerably in excess of 
full sunlight infrared radiation, in- 
dicating that heating effects are not re- 
sponsible for the folding response. Illu- 
mination of the pulvinus, or of whole 
leaves, with light in the waveband 
400-500 nm at a quantum fluency rate 
equivalent to that present in full sun- 
light (in the same waveband) was as 
effective in causing leaf folding as 
unfiltered sunlight, indicating that this 
portion of the spectrum is solely re- 
sponsible for the leaf-folding response. 
Finally, determinations of the effect 
of monochromatic light were also made 
in the range of 375-740 nm at approx- 
imately 25-nm intervals. (Shorter inter- 
vals were used at the wavelengths in 
the range 300-540 nm.) As expected, 
no action could be detected at wave- 



360 380 

500 520 540 560 



Fig. 35. Action spectrum for the rapid folding response of Oxalis oregana leaves. Prior to each 
determination, the petiole of each leaf was excised under water and placed in a small vial in the 
laboratory under illumination with white light at a quantum fluence rate of 50 ^mol m~ 2 s _1 for at 
least 30 min. After the leaflets had assumed an essentially horizontal position, a beam of 
monochromatic light (half-bandwidth = 12-15 nm) at a quantum fluence rate of 150 ^imol m -2 s _1 
was focused on a 4-cm 2 area centered on the junction of the three pulvini, and the change in leaflet 
angle {a) was determined for the next 10 min. The change in cosine a during the first 4 min was taken 
as a measure of the action. If no change in leaf angle could be detected in 10 min, the quantum fluence 
rate was raised to 300 yumol m -2 s _1 , and if there was still no change in leaf angle the action was con- 
sidered to be zero. Different leaf samples were used for each determination, and at least three 
replicates were used for determination of the action at each wavelength. 

lengths beyond 500 nm. An action 
spectrum for the folding response in 
the range 375-535 nm (Fig. 35) shows 
that radiation in the waveband 440- 
490 nm is most effective in bringing 
about the response. Some action was 
also obtained in the waveband 400-435 
nm. The abrupt decline in the action 
spectrum between 486 and 493 nm is es- 
pecially noteworthy. It suggests the 
possibility of an action on leaf unfolding 
that opposes the folding response and 
that is induced by longer wavelengths. 

The main features of the action spec- 
trum for leaf folding in Oxalis gener- 
ally resemble those of action spectra 
for a variety of phenomena involving 
blue-light-induced movements in alga, 
fungi, and higher plants. (Senger, 1980; 

Haupt and Feinleib, 1970). Although 
the action spectrum shown in Fig. 35 is 
not inconsistent with the notion that a 
flavoprotein serves as the photorecep- 
tor, no direct evidence concerning the 
identity of the photoreceptor is yet 

Implications of leaf folding in Oxalis 
leaves exposed to direct sunlight on 
photosynthetic function are discussed 
in the following article. 


Haupt, W., and M. E. Feinleib, eds., Encyclo- 
pedia of Plant Physiology, New Series, Vol. 7, 
Physiology of Movements, pp. 1-731, 
Springer, Berlin, 1979. 

Senger, H., ed., The Blue Light Syndrome, 
Springer, Berlin, Heidelberg, New York, 1980. 






Stephen B. Powles and Olle Bjorkman 

As is evident from the preceding 
report, the shade species Oxalis ore- 
gana (Nutt.) responds by folding down 
its leaves when suddenly exposed to in- 
tense light. In this article we will con- 
sider how this response affects the 
radiation environment of the leaves and 
the role it may serve in protecting the 
leaves from damage by excessive radia- 

To provide answers to these ques- 
tions, measurements of incident radia- 
tion, rate and extent of leaf movement, 
and the effect of such movement on 
photosynthetic performance and pho- 
toinhibition of Oxalis oregana leaves 
were conducted at different sites on the 
floor of a redwood forest. Complemen- 
tary studies were also made in the 
laboratory at Stanford. The field stud- 
ies were made during the period from 
March 30 to April 20, 1981, in the Wurr 
Flat area within the San Mateo 
Memorial Park, San Mateo County, 
California. We thank Mr. William 
Lawrence, Park Supervisor, for his in- 
terest and help in this project. The 
study period, characterized by clear 
skies, was preceded by heavy rainfall, 
ensuring excellent soil-water relations 
during our studies. 

Our mobile laboratory, used in these 
studies, in addition to its usual equip- 
ment for continuous gas-exchange anal- 
ysis, was equipped with instrumen- 
tation for micrometeorological and flu- 
orescence kinetic measurements. The 
latter instrument is similar in design to 
that described elsewhere in this Report 
(Fork, Oquist, and Powles), except that 
it is field-portable and housed in a 
special enclosure, thereby permitting 
us to maintain the leaf samples in total 
darkness for a period before and during 
the addition of liquid nitrogen to the 
leaf sample in the optical system. Flu- 

orescence emission was measured at 
690 nm; excitation was at 470 nm with 
a quantum fluence rate of 0.4 /*mol 
m -2 s _1 . 

For measurements of photosynthetic 
gas exchange in situ, a single leaf (three 
leaflets) with its petiole was enclosed in 
a specially designed chamber with a 
transparent, hemispherical top. This ar- 
rangement permits illumination of the 
leaf with a natural light field and allows 
free movement of the leaflets. Wire sup- 
ports were used to maintain the leaf in a 
horizontal position when desired. 

Studies were obtained from plants of 
Oxalis oregana in three native sites 
described below. Not shown are addi- 
tional studies of the daily carbon 
balance obtained from Oxalis at three 

Of the study sites, Site #1, in the deep 
shade, was never reached by suffi- 
ciently bright light to trigger a leaf- 
folding response or to supersaturate 
photosynthesis. The total daily quan- 
tum fluence rate at Site #1 was 0.73 mol 
m~ 2 day -1 and the peak fluence rate 
(averaged over 10 min) was 153 jumol 
m -2 s _1 . The corresponding values for 
Site #2, situated on the edge of an open- 
ing in the forest, were 2.3 mol quanta 
m~ 2 day -1 and 1590 /zmol quanta m -2 
s -1 . As much as 71% of the total daily 
quanta received at this site was con- 
tributed by two sunflecks, one just 
before and the other just after noon, the 
duration of the first and most intense 
sunfleck being about 20 min. Site #3, 
situated on the edge of a very large 
clear-cutting in the forest and facing 
southwest, permitted full insolation 
(about 1700 ^imol m -2 s _1 ) for periods 
of up to two hours. The total quanta 
received at Site #3 were 8.14 mol 
m -2 day -1 ; the major sunfleck, which 
arrived at this site at noon, contributed 



as much as 83% of the total. The Oxalis 
plants growing at Site #3 had a stunted 
appearance, and their chlorophyll con- 
tent per unit leaf area was lower than in 
plants growing at Sites #1 and #2. Site 
#3 is atypical for Oxalis oregana 
habitats and probably represents the 
extreme of radiation with which this 
species can cope. 

Figure 36 (upper) shows the effect of 
leaf folding on leaf inclination relative 
to a plane perpendicular to the sun's 
rays for the three leaflets of an Oxalis 
oregana plant growing at Site #2. The 
bottom part of Fig. 36 shows the actual 
quantum fluence rates incident on the 
leaflets. In this example, leaflet #3 had 
essentially the same azimuth as the sun 
(i.e., the axis from the pulvinus through 
the mid-rib of the leaflet was pointing 
south). Leaflets #1 and #2 each had an 
azimuth of approximately 120° in rela- 
tion to leaflet #3 (and the sun). At the 
instant the sunfleck struck the leaf 
(time = 0), the three leaflets were ori- 
ented in the same plane with an angle 
(i8) of 14° in relation to a plane perpen- 
dicular to the sun's rays. The solar 
radiation incident on the leaf was 97% 
(|8 = 14°; cosine /3 = 0.97) of the max- 
imum possible (/3 = 0, cosine & = 1.0). 
Within 20 s after the arrival of the 
sunfleck, the leaves started to fold 
down, and after approximately 4 min, 
leaflets #1 and #2 received only 18% 
(cosine fi = 0.18) of maximum fluence 
rate while leaflet #3 was still receiving 
about 56% (cosine (3 = 0.56). After 
about 6 min, leaflets #1 and #2 received 
no direct radiation (although they still 
received weak diffuse light) while 
leaflet #3 still received 30% direct light. 
Similar results were obtained with 
other Oxalis plants growing in Sites #2 
and #3, as well as in our laboratory 

The leaf-folding response has a signif- 
icant influence on the photosynthetic 
performance of Oxalis plants exposed 
to bright sunflecks in the field. In ex- 
periments conducted at Site #2, the rate 
of photosynthetic C0 2 uptake was 

measured in situ at a constant quantum 
fluence rate of 38 /xmol m -2 s _1 (using 
artificial light) before and after ex- 
posure to a natural sunfleck, having an 

lme, minutes 

Fig. 36. Time course of cosine ((3 = leaflet 
angle relative to a plane perpendicular to the 
sun's rays) (top) and of the actual quantum 
fluence rate (7) received by the leaflets (bottom) 
for an Oxalis oregana plant growing at Site #2. A 
sunfleck with a quantum fluence rate of 1590 
/anol quanta m~ 2 s _1 (horizontal plane) struck 
the leaf at time 0. The solar zenith (z) was 30°, 
the solar azimuth (s) was ±180° (south), and the 
leaf angle to horizontal (a) was 16° at time 0. The 
leaf azimuth (/) for leaflet #3 (A-A) was ±180° 
(same as the sun), and leaflets #1 and #2 (•-•) 
had azimuths of 60° and —60°, respectively. 
Cosine (3 was calculated from the expression 

cos /3 = cos a X cos z + sin a X sin z X cos (s — /). 

The quantum fluence rate (direct radiation only) 
received by the leaflets \I) is given by the expres- 
sion I = (IJcos z) X cos (3, where I is the quan- 
tum fluence rate incident on a horizontal plane. 
The dotted line depicts the quantum fluence rate 
that would have been received by the three leaf- 
lets if they had maintained their original posi- 
tion (a = 16°). 



average quantum fluence rate of 1500 
/imol m -2 s _1 and a duration of 18 min. 
(The leaf temperature was prevented 
from rising during the exposure.) When 
the leaflets were held in their origi- 
nal (essentially horizontal) position 
throughout the experiment, the rate at 
38 ptmol quanta m -2 s _1 measured after 
the departure of the sunfleck was 30% 
lower than before the arrival of the 
sunfleck. (The rate returned to the pre- 
sunfleck value over a 2-hour period.) In 
a similar experiment on the same plant, 
where the leaflets were permitted to 
fold down naturally, no such inhibition 
could be detected. The high-light-in- 
duced inhibition of photosynthetic rate 
was not caused by partial stomatal 
closure, since the stomatal conductance 
and the intercellular C0 2 pressure, after 
the exposure to the sunfleck, were 
higher than before the exposure. Thus, 
the decline in photosynthetic rate must 
reflect an inhibition of intrinsic photo- 
synthetic reactions. 

It is noteworthy that the photosyn- 
thetic rate during the sunfleck was as 
high when the leaf was permitted to 
fold down as when it was held in a 
horizontal position. This result is un- 
derstandable in view of the fact that 
light saturation of photosynthesis is 
reached at quite a low quantum fluence 
rate in this shade species. Our measure- 
ments show that for the Oxalis oregana 
plants at the three experimental sites 
the quantum fluence rates required to 
reach 50% and 90% of the light-sat- 
urated photosynthetic rate (3.0 jmiol 
C0 2 m~ 2 s -1 ) were about 30 and 100 
^mol quanta m -2 s _1 , respectively. The 
light intercepted by the leaves during 
the sunfleck must obviously have ex- 
ceeded 100 /xmol m -2 s _1 even when 
they were in the folded position. The 
results reported in the preceding article 
indicate that at incident quantum 
fluence rates in the range of approx- 
imately from 300 to 1200 /miol m~ 2 s _1 , 
the leaf-folding response is not a simple 
"on/off" switch. Rather it may serve to 
adjust the leaf angle so that the light 

intercepted by the leaves is maintained 
at sufficiently high levels to permit 
photosynthesis to operate at light sat- 
uration while at the same time avoiding 
excessive light levels that would cause 
injury to the photosynthetic system. 

Measurements of fluorescence at liq- 
uid nitrogen temperatures show that a 
substantial decline in the variable com- 
ponent, F v , takes place when Oxalis 
leaves are exposed to bright light 
(Tables 9 and 10). As discussed else- 
where in this Report (Fork, Oquist, and 
Powles; Bjorkman, Powles, Fork, and 
Oquist), such specific quenching of F v 
of photosystem II with little or no ef- 

TABLE 9. Redaction in Variable Fluorescence 
After Exposure of Oxalis oregana Leaves to In- 
tense Sunflecks in Native Sites #2 and #3* 

Reduction in F v 


Initial Value^ 

Most Other 

Exposed Portions 


Portion of the 

for Whole 


of Leaf Leaf 


A. Leaves 

held horizontal 



















Mean ± S.D. 

47 ± 9 

47 ± 9 

B. Leaves 

allowed to fold 


24 (13) t+ 



11 4 



26 12 



33 (14) +t 



31 11 


Mean ± S.D. 

25 ± 9 ± 12 

9 ± 5 

♦Variable fluorescence is defined as F v = 
(F m — F )IF o . The mean quantum fluence rates 
and duration of exposure for Site #2 were 1573 ± 
158 /umol m -2 s _1 and 20 min, respectively. 
Corresponding values for Site #3 were 1638 ± 59 
^mol m~ 2 s _1 and 50 min. 

+ The initial value for F v at the two sites was 
5.06 ± 0.75. 

^Values in parenthesis indicate apparent 
stimulation; these values are assumed to be zero 
in calculations of the average reduction in F v for 
the whole area of the leaf. 



TABLE 10. Effect of Exposure Time and 
Quantum Fluence Rate on Subsequent Reduc- 
tion in Variable Fluorescence (F u )* in Oxalis 
oregana Leaves ^ 


in F,„ 

% of Initial* " 





Rate, /^mol 



Free to 

— 2 —1 

m ^ s 





































*(F V before exp. — F v after exp.)/F ( , before 
exp.; mean value of F v before exposure was 6.20 
± 0.49 (n = 14). 

'The light beam, from a Metal-arc lamp, was 
vertical in these laboratory experiments. 

feet on F may be attributed to photoin- 
hibitory inactivation of the photosys- 
tem II reaction centers. 

As shown in Table 9, exposure of Ox- 
alis leaves growing at Sites #2 and #3 to 
intense sunflecks while preventing the 
leaves from folding down, resulted in 
47% average reduction of F v . The 
reduction was much smaller in ex- 
periments where the leaves were free to 
move. Significant reduction could only 
be detected in that portion of the leaf 
area having an azimuth similar to that 
of the sun; thus it was exposed to 
moderately bright light in spite of the 
folding response. In leaves permitted to 
fold, the average reduction in F v over 
the entire area of the leaf was about 9%. 

Table 10 shows the time course for 
the reduction in F v in Oxalis oregana 
leaves exposed to two different quan- 
tum fluence rates in the laboratory. In 

this experiment, the light beam was 
vertical and the leaves were held in a 
horizontal position. It is evident from 
these results that the extent of the 
decline in F v increases with both the 
duration and the intensity of the ex- 
posure. There was only a small effect on 
F v when the leaves were free to move 
under these conditions when the light 
beam was vertical. 

It should be noted that high-light- 
induced reduction in F v tends to be 
more pronounced than the reduction in 
photosynthetic rate. This may be ex- 
pected, since most of the fluorescence 
signal comes from those chloroplast 
layers in the leaf that receive the most 
intense light during the sunfleck. The 
photosynthetic rate represents the in- 
tegrated activity of all chloroplasts in 
the leaf. 

It is noteworthy that leaves of 
another shade species, Trillium ova- 
turn, growing adjacent to Oxalis ore- 
gana at Site #3, suffered a reduction in 
F v following exposure to the intense 
sunfleck at this site. This reduction in 
F v in Trillium (which does not possess a 
leaf- folding response) was about 40%, 
comparable to that of Oxalis plants 
that were prevented from folding their 

The leaf-folding response of Oxalis 
oregana thus provides an important 
mechanism for minimizing pho- 
toinhibitory injury to the photosyn- 
thetic system in this shade species. It 
is likely that the leaf-folding response 
has the additional advantage of improv- 
ing the water and heat balance of the 
leaves during exposure to high radia- 
tion. One may expect this latter func- 
tion to be particularly important later 
in the season when soil-water relations 
are less favorable, temperatures are 
higher, and the evaporative demand is 







Jeffrey R. Seemann, James M. Tepperman, and Joseph A. Berry 

Last year (Year Book 79, 146- 
147), we reported a survey of rates of 
photosynthetic C0 2 fixation by species 
of winter annuals native to and growing 
in Death Valley, California. While con- 
siderable variation was found, the 
photosynthetic capacities of these C 3 
species were in general much higher 
than those of other species having this 
pathway of C0 2 fixation, and were 
similar to those of the most productive 
C 4 species. The original report (Mooney 
et al, 1976) of a high photosynthetic 
rate (-60 /xmol m~ 2 s _1 ) in an in- 
dividual of the C 3 species Camissonia 
claviformis suggested that such capac- 
ity could be the result of a combination 
of high concentrations of the C0 2 -fixing 
enzyme ribulose-l,5-bisphosphate 
carboxylase-oxygenase (Rubisco) and 
of low diffusional resistances to the 
movement of C0 2 through the sto- 
mates. We have examined some of 
the physiological and biochemical char- 
acteristics of some of the more com- 
mon desert winter annuals in an at- 
tempt to account both for these high 
capacities for carbon fixation and for 
apparent species-specific differences 
observed in the photosynthetic capac- 
ity per unit of leaf-protein nitrogen. Our 
results indicate that these high 
capacities for photosynthesis are not 
the result of any apparent differences 
in the basic processes of C 3 photosyn- 
thesis, but rather can be at least par- 
tially explained by levels of leaf pro- 
tein, particularly Rubisco, that are 
significantly higher than in many C 3 
species from a wide range of other en- 
vironments (Bjorkman, 1980). Further- 
more, leaf internal C0 2 concentrations 
consistent with and necessary for this 
photosynthetic performance have been 
demonstrated. Most notably, however, 

differences in photosynthetic capacity 
between two of these desert winter an- 
nual species appears to be the result of 
differences in the kinetic properties of 

Materials and Methods 

Determinations of photosynthetic 
rates and stomatal conductances were 
made on plants occurring naturally in 
Death Valley, California (for details, see 
Year Book 79, 146-147). Leaves used 
for these measurements were frozen in 
liquid nitrogen for later biochemical 
analysis in the laboratory. Plants for 
biochemical studies were also grown in 
controlled-environment chambers at 
Stanford. Rubisco concentrations were 
determined quantitatively and specifi- 
cally by the radioimmuno-precipitation 
technique of Collatz et al. (Year Book 
78, 171-175). Total protein concentra- 
tions were determined by Kjeldahl 
analysis. Soluble protein concentra- 
tions were determined by Coomasie 
Blue protein binding (Bradford, 1976) 
after calibration against known concen- 
trations of bovine serum albumin 
(BSA), spinach soluble protein, and 
Rubisco. It should be noted that 
Rubisco displayed an approximately 
twofold higher reactivity per unit pro- 
tein than did BSA in this assay. 

Kinetic studies of Rubisco were per- 
formed essentially as described by 
Lorimer et al. (1977), either with crude 
leaf extracts or with purified enzyme 
prepared immediately prior to experi- 

Results and Discussion 

The levels and relative allocational 
patterns of total protein, soluble pro- 



tein, and Rubisco for nine species of 
desert winter annuals are shown in 
Figs. 37 and 38. Despite large differ- 
ences in absolute levels of proteins, the 
allocational patterns between protein 
pools for these species appears quite 
consistent. Furthermore, these protein 
ratios are quite typical of C 3 plants in 
general. The protein concentrations are 
usually quite high, with Rubisco con- 
centrations as high as 500 ng cm -2 in 
some field-grown individuals. Appar- 
ently, these annual species possess the 
capacity to accumulate large quantities 
of nitrogen from their native environ- 
ment for the production of photosyn- 
thetic machinery, most notably 
Rubisco. Certainly other components 
of the photosynthetic apparatus in ad- 
dition to Calvin cycle enzymes must be 
concomitantly increased (i.e., chloro- 
phyll concentration, which is maintained 
in a ratio with Rubisco of ~ 1:6 [g/g] in 

these species, typical of C 3 species ex- 
amined to date [Berry and Downton, 
1981]), and we have seen no evidence 
suggesting mechanistic differences be- 
tween the photosynthetic functioning of 
these species and that of other C 3 

Accompanying this enhanced car- 
boxylation capacity is the necessity for 
the maintenance of a sufficient internal 
C0 2 concentration to make effective 
use of such high Rubisco concentra- 
tions. Our previous study (Year Book 
79, 146-147) documented the existence 
of high stomatal conductances to C0 2 
in these species. Calculation of the ratio 
of internal C0 2 concentration (C £ ) to 
ambient C0 2 (C ), a measure of the ex- 
tent of nonstomatal limitations on pho- 
tosynthesis (Berry and Downton, 
1981), reveals that the value of CyC 
(0.73 ± 0.04) for these desert winter an- 
nuals is similar to values reported for 







1000 - 

500 - 

Death Valley annuals 

r=0. 89 

A A 




is. & A RuBP carboxylase 
'\ r=0.85 







Total protein, ^g c 



Fig. 37. The relationship between either soluble protein (A) or Rubisco concentration (•) and 
total protein concentration for nine species of desert winter annuals growing in Death Valley, 
California. The average Rubisco concentration is 18% of total protein, while soluble protein is 50% of 
total protein. 






d) 400 


">L 300 


L 200 



1 1 ' 1 ' 1 

Death Valley annuals 

• % 



.* V/^' 

• *y + 

• / * 

- /^ * 

• y/ 

!T •• 

— y** • 


-/ r=0. 91 


I 1 I 1 I 1 



Soluble protein, pa ci 


Fig. 38. The relationship between Rubisco 
concentration and soluble protein concentration 
for the same individuals as in Fig. 37. Rubisco 
represents 36% of soluble protein in these 

other C 3 species, and is independent of 
the rates of photosynthesis (Fig. 39). 
By comparison, the value of Q/C for C 4 
species is typically 0.2-0.4, a conse- 
quence of the C0 2 -concentrating func- 
tion of this pathway. Photosynthetic 
capacity of these C 3 species would be 
strongly reduced if stomatal limitation 
of C0 2 uptake were comparable to that 
of C 4 species. The conclusion to be 
drawn is that the very high photosyn- 
thetic capacities of these C 3 desert win- 
ter annuals, while comparable to rates 
in agriculturally important and native 
C 4 species, are obtained by quantita- 
tive increases in the capacity of steps of 
C 3 photosynthesis rather than by qual- 
itative modification of carbon-fixation 

It was also noted earlier (Year Book 
79, 146-147) that two closely related 
species of desert winter annuals, 
Camissonia claviformis and C brevipes 
(Onagraceae), possessed rates of photo- 
synthesis that were, on average, signif- 
icantly higher than those of all other 
species. This difference could not be ac- 
counted for simply on the basis of pro- 
portionately higher concentrations of 

Rubisco or associated proteins. Indi- 
viduals of Geraea canescens (Aster- 
aceae), another common desert annual, 
have been determined to contain Ru- 
bisco in concentrations equal to Cam- 
issonia, yet possess lower photosynthe- 
tic capacities. This observed difference 
in the apparent efficiency of Rubisco 
between these two genera in relation to 
photosynthetic capacity is shown in 
Table 11 (right column). For equal 
Rubisco concentrations, the whole-leaf 
C0 2 fixation capacity of Camissonia 
exceeds that of Geraea by ~ 1.8 X. This 
difference is uncorrected for differences 
in C t between species. As is shown below, 
however, the existence of such a dif- 
ference (values of C, for Camissonia 
were -1.07X greater than those of 
Geraea) cannot account for the entire 
difference in the efficiency of Rubisco 

To provide at least a partial explana- 
tion for this phenomenon, we have ex- 
amined the in vitro kinetic properties of 
Rubisco from these two genera. Differ- 
ences in the specific activity (fimol 




1 1 



3 y 

I ' 



1 1 













• •• • 


t , 


• • 

























i 1 









Fig. 39. The rate of whole-leaf photosyn- 
thesis for nine species of desert winter annuals 
as a function of the ratio of intercellular C0 2 
concentration to ambient C0 2 concentration. (C,~/ 
C can be shown to be mathematically equivalent 
to the ratio of the mesophyll resistance to the 
total [mesophyll -I- stomatal] resistance). 



TABLE 11. Values for the Kinetic Efficiency of Rubisco from Two Species of 
Desert Winter Annuals* 

Kinetic constants in vitro at 25 °C 

Photosynthetic Efficiency 
/xmol CO 2 min -1 mg Rubisco" 


(^mol CO 2 min 
mg Rubisco ) 

K m (C0 2 ) 

(Mir 1 ) 



Camissonia sp. 

2.85 ± .26 

310 ± 21 

0.77 ± .18 

0.73 ± .19 

(n = 15) 

(n = 5) 

in = 12) 

Geraea canescens 

2.40 ± .41 

331 ± 50 

0.51 ± .09 

0.41 ± .04 

in = 14) 

(n = 4) 

(n = 6) 

*In vitro kinetic constants were determined on the basis of 14 C0 2 incorporation into acid-stable 
products. Photosynthetic efficiency was determined on the basis of gas-exchange analysis of whole 
leaves. Data for Camissonia brevipes and C. claviformis is placed together as Camissonia sp. 
Predicted photosynthetic efficiency was calculated as described in the text using kinetic constants 
determined in vitro for Rubisco, measured stomatal conductances and a leaf temperature of 20 °C. 
Values of V max and K m (C0 2 ) were calculated by the method of Wilkinson (1961). 

14 C0 2 fixed • min -1 • mg Rubisco -1 ) 
(measured at saturating HC0 3 ~ and 
RuBP concentrations) of the enzyme 
between Camissonia and Geraea were 
found to parallel the previously dis- 
cussed differences in the apparent effi- 
ciency of Rubisco determined on the 
basis of whole-leaf photosynthesis. 
Camissonia Rubisco was found to have 
a specific activity -1.2X higher than 
that of Geraea. (See Table 1 1 for actual 
in vitro values.) We have found that 
values for Rubisco specific activity are 
reproducible only within a certain 
range, as the enzyme in our hands suf- 
fers a nonrecoverable loss of activity 
within the first hour after purification. 
Variability in the maximal state of acti- 
vation of the enzyme, despite the pres- 
ence of effectors such as 6-phospho- 
gluconate, also contributes to a degree 
of noise in our specific activity determi- 
nations. Determination of the Michaelis 
constants for C0 2 (K m (C0 2 )) for Ru- 
bisco from these species has revealed 
no significant difference between 
genera, as is shown in Table 11. The 
values of -310-331 jxl/liter (-10 /JVI) 
are in agreement with previously re- 
ported values in the literature for the 
fully activated enzyme. 

In order to evaluate our results in the 
light of present-day knowledge of the 

kinetic behavior of Rubisco in the 
photosynthetic process, we have used a 
biochemical model of C 3 photosyn- 
thesis based upon the Michaelis- 
Menten equations for enzyme catalysis 
in the presence of a competitive in- 
hibitor. This allows us to predict the 
whole-leaf rate of photosynthesis given 
our measured values of Rubisco V max , 
K m (C0 2 ), C0 2 concentration, and 
stomatal conductance (in order to 
calculate the effective internal C0 2 con- 
centration). For our purposes here, we 
have assumed that both the carboxyla- 
tion and oxygenation reactions of 
Rubisco are RuBP-saturated. This 
assumption is probably valid for the 
light and temperature regimes used 
during the measurements of photosyn- 
thesis considered here. This model 
allows us to determine on the basis of 
Michaelis-Menten kinetics if our 
biochemical measurements can account 
for both the relative differences ob- 
served in photosynthetic performance 
of Camissonia and Geraea and the ab- 
solute magnitude of their photosyn- 
thetic rates. Figure 40A demonstrates 
that if the model is used to predict 
photosynthetic rates using the specific 
activity of spinach Rubisco as deter- 
mined by Badger (1976) (2.2 /xmol/ 
min/mg Rubisco at 25 °C) for both Ca- 






















40 - 


20 - 




> 1 i 1 i 

specific activities 












• J 






• . /* 


Camissonia - 




Geraea ~ • 




r=0. 96 

1 1 1 1 1 





1 — i — ' — i — ' — i — ' — r~ 

Measured specific activities ** 

Camissonia —& 
Geraea " • 


40 50 10 20 

Observed Ps rate 



C0 2 m 

-2 -1 

Fig. 40. (A) Model prediction of the whole-leaf rate of photosynthesis for Geraea and Camissonia 
and comparison to rates actually observed, assuming in the model that Rubisco kinetic properties 
for these two species are equivalent (see text for details). (B) Model prediction and comparison to 
observed rates, assuming kinetic properties of Rubisco for these species as were determined in vitro 
(see Table 11 and text). Solid lines show linear regressions of the data, and a perfect agreement be- 
tween the predicted and observed relationships are shown by the dotted lines. 

missonia and Geraea, there is not ex- 
act agreement with the observed rates 
of photosynthesis. However, if our 
measured in vitro values for these two 
species (Table 11) are substituted in the 
model, the agreement between ob- 
served and predicted rates of photosyn- 
thesis for both species is improved (Fig. 
40B), particularly in the case of 
Camissonia. These data are also sum- 
marized in Table 11, where on the basis 
of predicted photosynthesis per unit of 
Rubisco, it can be seen that a difference 
of —1.5X exists between Camissonia 
and Geraea (after adjustment for C,- dif- 
ferences), in close agreement with the 
observed values (Table 11). 

That we can account for the rates of 
C0 2 fixation observed under these con- 
ditions indicates that our techniques for 
extraction, quantification, and kinetic 
analysis of Rubisco at least predict suf- 
ficient biochemical capacity. Further- 
more, these techniques provide a plausi- 
ble explanation for apparent differences 
between Camissonia and Geraea at a 

physiological level based upon differ- 
ences in the properties of Rubisco mea- 
sured in vitro. We are confident that 
there is a significant difference in the 
F max of Rubisco of these two species. 
However, other factors not yet consid- 
ered in our study may affect photosyn- 
thetic performance in vivo. Most criti- 
cally, we know very little concerning the 
control of the activation state of 
Rubisco (in vivo), and we have not 
tested the assumption that RuBP levels 
were saturating for C0 2 uptake under 
the conditions used for photosynthetic 
measurements in vivo. It is also conceiv- 
able that the concentration of C0 2 at the 
site of its fixation in the chloroplast 
stroma may differ significantly from 
that in the intercellular air spaces. Fur- 
ther experiments are under way to ex- 
amine these possibilities and to qualify 
more precisely the kinetic properties of 
Rubisco from these and other species. 
How widespread variation in Rubisco 
kinetic properties will turn out to be re- 
mains an unanswered question, as evi- 



dence for variability of this enzyme is 
decidedly limited (for review, see Ogren 
and Hunt, 1978). In conclusion it should 
be emphasized that these studies have 
considered the relationship between bio- 
chemical characteristics and whole-leaf 
photosynthesis studied under a single 
condition chosen such that Rubisco ac- 
tivity would likely be rate-limiting for 


Badger, M. R., Ph. D. Thesis, Aust. Nat. Univ., 

Berry, J. A., and W. J. S. Downton, CO2 fixation 
and plant productivity, in Photosynthesis: 
Volume 2, Govindjee, ed., in press, 1981. 

Bjorkman, O., in Encyclopedia of Plant Phys- 
iology, New Series, Vol. 12, O. Lange, ed., in 
press. 1980. 

Bradford, M. M., Anal. Biochem., 72, 248-254, 

Lorimer, G. H., Annu. Rev. Plant Physiol., 32, 
349-383, 1981. 

Lorimer, G. H., M. R. Badger, and T. J. An- 
drews, Anal. Biochem., 78, 66-75, 1977. 

Mooney, H. A., J. Ehleringer, and J. A. Berry, 
Science, 194, 322-324, 1976. 

Ogren, W. L., and L. D. Hunt, in Photosynthetic 
Carbon Assimilation, 179-208, H. W. Siegel- 
man and G. Hind, eds., Plenum Press, 1978. 

Wilkinson, G. N., Biochem. J., 80, 324-332, 1961. 

Lavatera cretica AND Malva parviflora 

Dov Roller 

Schwartz and Koller (1978) have 
shown that the photoreceptor for the 
vectorial phototropic response of 
Lavatera cretica and Malva parviflora 
leaves is located in the lamina, prob- 
ably in association with the veins, not 
the mesophyll or pulvinus. They have 
suggested that the photoreceptor is ex- 
cited by oblique illumination of a vein 
in its plane of symmetry. The nature of 
the response depends on the direction 
of the oblique beam. A tip-oriented 
beam causes the vein to incline (up- 
ward) by transmitting a signal that 
results in an increase in turgor in the 
segment of the pulvinar motor tissue 
associated with this vein, while a 
similar, but base-oriented beam causes 
the vein to decline (downward) by 
decreasing turgor in the same segment. 

The exact location and biochemical 
nature of the photoreceptor are un- 
known, except that it is excited by the 
blue, not red, region of visible radiation. 
Its physical organization within the cell 
is fundamental to its mode of action. 
The photoreceptor molecules could all 
be similarly oriented in a unique 
geometrical relationship to the axis of 
the vein, or they could be attached in a 

nonspecific arrangement on the surface 
of some optically dispersive or opaque 
subcellular structure situated at right 
angles to the axis of the vein. If the 
molecules are uniquely oriented, one 
can expect the response to be strongly 
dependent on angle of incidence of the 
oblique beam on the vein. On the other 
hand, in the case of the alternative ar- 
rangement any oblique illumination of 
the linear array of subcellular struc- 
tures should result in different fluences 
reaching the photoreceptor molecules 
on their opposite sides, with relatively 
small effect of angle of incidence 
(smaller than the critical). One way of 
discriminating between these alter- 
native possibilities is to establish the 
quantitative relationship between an- 
gle of incidence and kinetic parame- 
ters of the response. This, however, is 
not a simple undertaking, since as far as 
we know the response to light is non- 
inductive. This means that if the obli- 
que beam is fixed, its angle of incidence 
on the lamina would change pro- 
gressively as the lamina reorients, until 
it becomes normal to the beam. Such a 
situation evidently does not occur dur- 
ing actual solar tracking, where the 



solar azimuth is continuously chang- 
ing. Therefore, a way had to be found to 
determine the kinetics of laminar 
response while maintaining a constant 
angle of incidence. 

One approach was based on the 
possibility that the response might be 
studied in detached leaves, as was the 
case in leaves exhibiting photonastic 
and circadian movements (Hillman and 
Koukkari, 1967; Satter et al., 1974a, b), 
or even in some leaves exhibiting 
phototropic responses (Wien and 
Wallace, 1973). To do this, laminas of 
detached leaves were placed flat on a 
horizontal perforated partition, through 
which their petioles were threaded into 
a water-filled vessel with transparent 
walls. The lamina thus remained sta- 
tionary, maintaining a constant angle 
of incidence with a fixed oblique beam 
directed at it. Phototropic response at 
the pulvinus could then only take place 
by movement of the suspended petiole, 
a process that could be continuously 
observed through the transparent walls 
using nonphototropic illumination if 
necessary. However, the water rela- 
tions of the leaf proved very sensitive 
to excision (although carried out under 
water, with or without the node), and 
led to wilting in light unless the leaves 
were allowed to recover in darkness for 
a number of hours. By that time, the 
submerged petioles were exhibiting 
curvatures unrelated to phototropism. 
These undesirable effects were not 
remedied by controlling the pH, by 
employing a nutrient solution, or by 
supplying sucrose. This approach was 
therefore abandoned as unreliable. 

A different approach was to tilt in- 
tact potted plants so as to bring a spe- 
cific leaf in each of them to the desired 
angle of incidence with a vertical beam, 
and keep on tilting as to correct for the 
laminar reorientation and keep con- 
stant the angle of incidence. The rate of 
tilting of each pot could be continu- 
ously monitored. As the plants used 
were in the rosette stage, it was hoped 
that geotropic responses of the short 

stem, if any, would not interfere with 
the phototropic response. However, it 
was found that the geotropic stimulus 
was either sensed by the petioles or 
transmitted to them, causing in them a 
rapid negative geotropic curvature 
even in darkness. This approach was 
therefore also abandoned. 

The limitations in the two previous 
approaches led to the present method, 
which employs intact, erect plants. By 
this method, the light source is con- 
tinuously reoriented on the lamina so as 
to maintain the angle of incidence con- 
stant despite laminar movement. The 
modified photoelectric device for record- 
ing leaf movements, developed by Tib- 
bitts et al. (1973), was initially used. 
However, it was designed for a type of 
leaf quite different from the malva- 
ceous, and it would have required ex- 
tensive structural modifications for 
this purpose. Hence, a manually opera- 
ted apparatus was designed and con- 
structed, in which the pulvinus of the 
treated leaf is maintained in a fixed 
position, and changes in the angle (in- 
clination or declination) of a single vein 
around this axis can be continuously 
followed with a resolution of 0.5°, with- 
out touching the leaf. The low infrared 
light source (Sylvania Cool-Lux Spot 
Lamp 150 PAR 38/2 SP) is directed to- 
ward the laminar surface at the desired 
angle of incidence, and its movement is 
manually linked to that of the lamina. 
This lamp is equipped with a dichroic 
reflector and can provide fluence rates 
up to 1000 /xmol m -2 s _1 between 400 
and 700 nm uniformly on the leaf sur- 
face, without appreciably overheating 
the leaf over extended periods. The 
plants are kept adequately watered 
and are ventilated by means of a fan. 
On the basis of extensive testing and 
modification of the prototype, addi- 
tional devices are under construction 
and nearing completion. These will 
enable replication and averaging of re- 
sults of simultaneous and equal treat- 
ment, as well as simultaneous compari- 
son of different treatments. 



During its period of testing, the pro- 
totype has already yielded some signifi- 
cant observations. For instance, at a 
given angle of incidence, the lamina can 
sustain a reorientation at a constant 
angular velocity for considerable 
lengths of time (about 1 h), before 
changing the rate (usually to a slower 
one). A stationary phase of variable 
duration (and of unknown origin) may 
occur, most commonly in the transition 
to a new angular velocity. In suitable 
leaves, angular velocity may readily ex- 
ceed that of the earth's rotation (i.e., the 
sun's azimuth), even though irradiance 
is only a fraction of the sun's. Maximal 
rates obtained were about 90° h _1 
(inclination or declination). Vectorial 
excitation of the lamina is transmissi- 
ble to the petiole (which is shaded 
against direct phototropic excitation), 
causing it to bend toward the beam, 
thus enhancing the reorientation by the 
pulvinar response. A similar transmis- 
sion was already reported from vec- 
torially excited cotyledons to the 
darkened hypocotyl (Schwartz and 
Roller, 1980). The extent to which the 
petiole participates in the vectorial 
phototropic response depends on its 
mechanical resistance to bending and is 
thus affected by plant age, leaf age, and 
growing conditions. Whereas pulvinar 
bending results from differential, rever- 
sible changes in cell volume, bending of 
the petiole may or may not involve dif- 
ferential and irreversible extension 
growth. Observations on solar-tracking 
leaves of other species, which lack mor- 
phologically distinct pulvini, suggest 
that the entire petiole may be operating 
as a "diffuse" pulvinus. 

It was also found that once laminar 
reorientation (declination) had pro- 
ceeded for some time under constant 
base-oriented excitation, it could con- 
tinue at virtually the same rate for a 
considerable period (1 h or longer) after 
the light had been turned off. Reex- 
amination of previous data (obtained 
with a fixed beam) indicated that a 
similar phenomenon may take place 

also in the tip-oriented response. Con- 
tinued reorientation in darkness follow- 
ing vectorial excitation could be a 
result of phototropic induction or of in- 
ertial translocation of solute and water 
that had started during vectorial ex- 
citation. This phenomenon merits fur- 
ther study, as well as verification for 
the tip-oriented response, using the 
leaf-tracking beam. 

Previous studies showed that when a 
lamina which had inclined upwards in 
response to tip-oriented vectorial ex- 
citation is transferred to darkness, it 
starts to decline and comes to rest at 
approximately 90° to the petiole. This 
is the typical "night" position (until 
onset of the predawn reorientation to 
face sunrise). Base-oriented excitation 
of such inclined leaves causes the 
lamina to reorient in the same direction 
but at a faster rate. Furthermore, the 
response to a tip-oriented fixed beam 
increased linearly with log fluence rate 
(up to about 400 ptmol m~ 2 s _1 ), while the 
response to a base-oriented fixed beam 
was saturated at a very low level (about 
30 /miol m~ 2 s _1 ). These results sug- 
gest that (1) excitation by a tip-oriented 
beam may lead to active uptake of 
solutes into the motor cells, (2) a return 
to darkness stops this uptake, leading 
to passive dissipation of the osmotic 
potential gradient, and (3) excitation by 
a base-oriented beam accelerates the 
passive dissipation, possibly by in- 
creasing membrane permeability to the 
solutes. However, by using a base- 
oriented, leaf-tracking beam to elicit 
the declination response, it was possi- 
ble to cause the lamina to decline 
almost vertically down, assuming quite 
acute angles (<90°) with the petiole, 
i.e., well beyond the 90° dark- 
equilibrium position. This suggested 
that the response to excitation by a 
base-oriented beam may result from ac- 
tive transport of solutes out of the 
motor cells. 

These nearly round malvaceous 
leaves are equipped with seven major 
veins radiating from the pulvinus (ap- 



proximately 40°, 85°, and 120° on 
either side of the midrib/central vein). 
It is therefore conceivable that a base- 
oriented beam for any single vein may 
simultaneously provide partial tip- 
oriented excitation for at least two 
other veins on the opposite side of the 
pulvinus. The active uptake on the op- 
posite of the pulvinus could thus sup- 
plement the leakage caused by the 
base-oriented excitation, leading to the 
extension of reorientation beyond the 
equilibrium position. 

This possibility was studied by using 
dissected leaves, possessing only the 
midvein and its immediate neighbor on 
either side, plus the connecting mes- 
ophyll, where the possibility of inadver- 
tent tip-orientation is eliminated. The 
cut surfaces were immediately sealed 
with lanolin to prevent damage by 
desiccation. Such leaves exhibited the 
full response to both tip-oriented and 
base-oriented excitation over several 
cycles, suggesting that both responses 
may operate by means of active 
transport, though in opposite direc- 

The photoreceptive tissues associ- 
ated with the veins still await iden- 
tification. Our attention was drawn to a 
peculiar strand of colorless collen- 
chymatous tissue overlying the veins 
and protruding well above the leaf sur- 
face. In preliminary studies, surgical in- 
terruption of the strand's continuity 
failed to affect the vectorial response, 
ruling out any role of it as an optical 
fiber. Surgical removal of long seg- 
ments (along the basal 1/2-2/3 of the 
major veins) did not have any apparent 

effect on the mechanical strength of the 
lamina, nor did it seem to have a signifi- 
cant effect on the capacity for the vec- 
torial phototropic response, even when 
the intact portions were shaded by a 
charcoal-lanolin paste. Again, the muti- 
lated portions of the leaf were pro- 
tected against desiccation by lanolin. 
Although these results cast doubt on 
participation of this tissue in photoper- 
ception, they are very preliminary and 
deserve further study. 


Dr. T. W. Tibbitts, Dept. of Horticul- 
ture, University of Wisconsin, Madi- 
son, provided loan of his apparatus; Dr. 
R. Stout, Dept. of Botany, University 
of Washington, Seattle, collaborated 
during a brief visit to this laboratory; 
Mr. R. Hart of this laboratory gave in- 
valuable assistance in the design and 
construction of the leaf-tracking appa- 


Hillman, W. S., and W. L. Koukkari, Plant 

Physiol., 42, 1413-1418, 1967. 
Satter, R. L., G. T. Geballe, and A. W. Galston, 

J. Gen. Physiol, 64, 413-430, 1974a. 
Satter, R. L., G. T. Geballe, and A. W. Galston, 

J. Gen. Physiol, 64, 431-442, 1974b. 
Schwartz, A., and D. Roller, Plant Physiol., 61, 

924-928, 1978. 
Schwartz, A., and D. Roller, Plant Physiol, 66, 

82-87, 1980. 
Tibbitts, T. W., G. D. Nutter, T. Hoshizaki, and 

R. R. Badeau, Plant Physiol, 51, 812-814, 

Wien, H. C, and D. H. Wallace, Crop Sci., 13, 

721-724, 1973. 




Michael G. Murray and William F. Thompson 

The major emphasis of the past year's 
research has been to determine if the 
general features of mRNA-coding DNA 
are similar to or different from total 
DNA. Two basic approaches have been 
taken: the first is a comparison of the 
pattern of repeat sequence intersper- 
sion in coding and noncoding regions of 
the pea genome. The second approach is 
to compare the chromatin organization 
of coding and noncoding DNA in pea 
and wheat. 

Pea Coding DNA Sequence Organiza- 
tion. Numerous studies on repeat se- 
quence interspersion have followed 
Britten and Davidson's (1969) hypothe- 
sis that interspersed repetitive ele- 
ments might play a role in the coordi- 
nate regulation of transcription. Data 
in sea urchin and mouse suggest that 
most transcribed sequences are con- 
tiguous with repetitive elements 
(Davidson et ah, 1975; Costantini et al, 
1980; Kuroiwa and Natori, 1979), and 
fine-scale analysis of specific genes has 
often revealed repetitive sequences in 
close proximity to transcribed DNA 
(e.g., Adams et al, 1980; Coggins et al, 
1980; Duncan et al, 1979). However, it 
is not yet entirely clear whether the oc- 
currence of repeats adjacent to mRNA- 
coding sequences reflects a functional 
relationship or merely the random dis- 
persal of repetitive elements. 

The pea genome is considerably larger 
and characterized by more extensive 
short-period repeat sequence intersper- 
sion than any of the animal systems 
used in the above studies. From pre- 
vious work, we know that 85% of 1300- 
nucleotide-long randomly sheared pea 
DNA fragments contain highly repeti- 
tive sequences (about 5000 copies per 
haploid genome) and essentially all of 
the rest contain repeats present in 
about 50 copies per haploid genome 
(Murray et al, 1978). It is also clear 

that the sequences transcribed into 
mRNA cannot represent more than 
1-2% of the pea genome (Murray et al, 
1981). Consequently, it was of interest 
to determine the extent to which the 
pattern of repeat sequence organiza- 
tion in the small fraction of the DNA 
that contains genes reflects that seen in 
the total genome (Murray and Thomp- 
son, in press). 

To address this question, randomly 
sheared pea DNA fragments 1300 nu- 
cleotides in length were separated on 
the basis of whether or not they contain 
highly repetitive elements by means of 
DNA reassociation and hydroxylapa- 
tite fractionation. The relative concen- 
tration of sequences complementary to 
mRNA in the fractions of long frag- 
ments that did or did not also contain 
highly repetitive elements were then 
compared by following the reassocia- 
tion of a cDNA tracer when driven by 
these two fractions as well as with total 
pea DNA. The mRNA-coding sequences 
were found to be enriched in the small 
fraction (about 10%) of 1300-nucleotide 
fragments that did not also contain 
highly repetitive sequences. They were 
found to be depleted in the fraction of 
long fragments bearing the high-fre- 
quency repeats. 

The fraction of long fragments en- 
riched for mRNA-coding sequences but 
depleted in high-frequency repeats was 
found to be composed primarily of low- 
frequency repeats and single-copy se- 
quences. Analysis of the thermal sta- 
bility of these low-frequency repeats 
indicated that they had undergone con- 
siderably less sequence divergence 
than the average pea repeat. The ex- 
perimental approach did not allow us to 
conclude whether or not the mRNA- 
coding sequences were present on the 
same 1300-nucleotide-long fragments 
as the low-frequency repeats. The data 



do however provide evidence that the 
genes occupy a significantly different 
domain of genome organization. 
Whereas most sequences in the pea 
genome are within 1300 nucleotides of 
a highly repetitive element, the genes 
themselves are not. In contrast, any re- 
peats that may he within a distance of 
1300 nucleotides of an mRNA-coding 
sequence must belong to a distinctive 
subset of pea repeats characterized by 
a relatively low copy number and an 
unusually low degree of sequence diver- 

To facilitate a more precise analysis 
of sequence organization in gene-con- 
taining regions, considerable effort was 
devoted to the isolation of pea DNA 
long fragments which contain tran- 
scribed sequences. These experiments 
were based on the observation that 
RNA:DNA hybrids are more stable 
than DNA duplexes in concentrated 
trichloroacetate salt solutions (Chien 
and Davidson, 1978). The optimum con- 
dition for the formation of RNA:DNA 
duplexes was found to be 45 °C in 3.0 M 
sodium trichloroacetate. Hybrid forma- 
tion proceeded at a rate equivalent to 
that observed in 0.12 M sodium phos- 
phate at 60 °C with little evidence of 
nucleic acid degradation, at least for a 
period of about 48 h. Since the T m for 
duplex DNA in 3.0 M sodium trichloro- 
acetate is about 32°C, DNA:DNA re- 
naturation could be completely elimi- 
nated. By treating the hybridization 
mixture with 1.0 M glyoxal (see Kaback 
et al, 1979) in 3.0 M sodium trichloro- 
acetate at room temperature, the non- 
hybridized DNA could be rendered non- 
renaturable, thereby permitting fur- 
ther manipulations in conventional 

Two approaches were used to isolate 
hybrid molecules. First, the RNA was 
mercurated prior to reassociation and 
the hybrids retained by sulfhydryl 
agarose affinity chromatography (Dale 
et al, 1975; Brown and Balmain, 1979). 
The second approach was to use poly-U 
sepharose to retain hybrid molecules 

involving polyadenylated RNA. While 
both approaches seemed feasible in 
model experiments, neither proved 
practical for the isolation of gene-con- 
taining fragments of total pea DNA. 
From 1 to 3% of starting 5000-10,000 
base pair fragments were typically re- 
tained upon the first cycle of fractiona- 
tion. While a fairly low fraction was ex- 
pected, it soon became obvious that 
any contamination with nonhybrid 
DNA would pose serious problems. It 
did not prove practical to prepare suf- 
ficient quantities of gene-containing 
fragments through several cycles of 
fractionation in order to allow further 
characterization. The general approach 
might be more useful in a smaller ge- 
nome such as the mung bean, where the 
fraction of coding DNA is expected to 
be much larger than in pea. 

Studies on Wheat and Pea Chroma- 
tin. Experiments are now in progress to 
compare the chromatin organization of 
coding and noncoding DNA in both pea 
and wheat. A vast body of evidence in 
animal systems suggests that tran- 
scriptionally active sequences are in an 
altered chromatin conformation. One of 
the major findings leading to this con- 
clusion is that potentially active genes 
are more sensitive to digestion by 
DNAses than are the bulk of the DNA 
in intact nuclei and chromatin. Very 
good correlations exist between the in- 
creased nuclease sensitivity and the 
preferential localization of a distinctive 
class of nonhistone chromosomal pro- 
teins called high-mobility-group pro- 
teins (reviewed in Mathis et al, 1980). 
To date, no comparable data exists for 
higher plants. 

Experiments done in collaboration 
with Dr. Steven Spiker at Oregon State 
University have examined whether 
transcriptionally active chromatin in 
wheat is preferentially sensitive to 
DNAse I. Following approaches devel- 
oped in animal systems (Mathis et al., 
1980), we isolated wheat chromatin or 
nuclei from unimbibed wheat germ and 
treated these preparations with DNAse 



I such that from 5% to 20% of the DNA 
was rendered acid soluble. We then 
compared the relative concentration of 
mRNA-complementary sequences in 
the DNAse-resistant DNA to that in 
total DNA by following the reassocia- 
tion of a cDNA tracer when driven with 
the various DNAs. cDNAs prepared 
from either dry embryo total poly- A + 
RNA or 3-h-imbibed polysomal poly-A + 
mRNA have been used with similar 

In initial experiments using wheat 
chromatin isolated according to Simon 
and Becker (1976), no preferential di- 
gestion of coding sequences was ob- 
served. However, this method involves 
the use of 50 mM ammonium sulfate 
and thus may have permitted some his- 
tone redistribution during chromatin 
isolation. When experiments were re- 
peated using chromatin isolated under 
low ionic strength conditions, prefer- 
ential digestion of mRNA-coding se- 
quences was observed. Preferential 
digestion was also seen when intact 
nuclei isolated under low ionic strength 
were used. Conditions that lead to the 
preferential digestion of transcription- 
ally active chromatin also lead to the 
release of high-mobility-group proteins. 

The data in wheat are thus consistent 
with those in various animal systems. 
An interesting difference in the wheat 
studies is that whereas all previous 
work in animals had been performed on 
transcriptionally active tissue, the 
studies in wheat were performed with 
unimbibed and thus transcriptionally 
quiescent embryos. It has now been 
well established that de novo RNA syn- 
thesis is essential for the rapid resump- 
tion of growth following imbibition 
(Jendrisak, 1980). The fact that the 
mRNA populations in dry wheat germ 
and in 6-h imbibed embryos are qualita- 
tively very similar (Caers et al, 1979) 
would be consistent with the view that 
the desiccated embryo is merely an ar- 
rested stage of embryogeny. One might 
logically expect that aspects of chroma- 
tin structure involved in gene regula- 

tion would be maintained in the desic- 
cated embryo. 

Differential nuclease sensitivity is 
also being used to examine pea chroma- 
tin, but with a somewhat different ap- 
proach. In various animal systems, 
specific sites hypersensitive to DNAse 
I have been identified in the vicinity of 
transcribed sequences. These sites are 
preferentially nicked with DNAse treat- 
ments that do not lead to any signifi- 
cant degradation of the DNA (e.g., 
Stalder et al, 1980). Experiments are 
under way to compare the relative sus- 
ceptibility to nicking for a variety of 
pea sequences: a transcribed tandem 
array as represented by ribosomal 
genes, a non transcribed tandem repeat 
as represented by a cloned repeat probe, 
and transcribed single-copy sequence. 

In these experiments, intact nuclei 
are isolated and samples treated with 
varying DNAse I concentrations to in- 
troduce nicks at a frequency ranging 
from 1 per 1000 base pairs to 1 per 
25,000 base pairs. The resulting DNA 
size distributions are resolved on low- 
percentage alkaline agarose gels, and 
the DNA is transferred to nitrocellu- 
lose (Southern, 1975). The nitrocellu- 
lose blots are then hybridized with 
various 32 P-labeled probes. The relative 
size of the complementary sequences 
represented by the probe can then be 
compared to that of the total DNA. 

Initial experiments on the ribosomal 
genes suggest that these genes are 
somewhat less subject to nicking then 
is the bulk of pea DNA. Leber and 
Hemleben (1979) reported that the ribo- 
somal genes in Brassica chromatin are 
less sensitive to digestion than bulk 
chromatin. While surprising, these re- 
sults may not be totally unexpected, 
for a number of reasons. If only some of 
the 8000-odd copies of the ribosomal 
gene in pea (Ingle and Sinclair, 1972) 
are actually transcribed at any one 
time, then one might not expect the 
total population to show increased 
nuclease sensitivity. The lower sen- 
sitivity to nicking observed for 



ribosomal genes relative to total DNA 
might be explained if in pea, as has 
been shown in corn, the majority of the 
ribosomal genes are heterochromatic 
(Givens and Phillips, 1976). Additional 
work is in progress to compare the 
relative frequency of nicking in se- 
quences transcribed into mRNA, as 
well as those in a cloned repeat showing 
a tandem pattern of organization. 


Adams, J. W., R. E. Kaufman, P. J. Kretschmer, 
M. Harrison, and A. W. Nienhuis, Nucl. Acids 
Res., 8, 6113-6128, 1980. 

Britten, R. J., and E. H. Davidson, Science, 165, 
349-357, 1969. 

Brown, T. D. K., and A. Balmain, Nucl. Acids 
Res., 7, 2357-2368, 1979. 

Caers, I. L., W. J. Peumans, and A. R. Carlier, 
Planta, 144, 491-496, 1979. 

Chien, Y. H., and N. Davidson, Nucl. Acid Res., 
5, 1627-1637, 1978. 

Coggins, L. W., G. J. Grindlay, J. K. Vass, S. S. 
Slater, P. Montague, M. A. Sinson, and J. 
Paul, Nucl. Acids Res., 8, 3319-3333, 1980. 

Costantini, F. D., R. J. Britten, and E. H. David- 
son, Nature, 287, 111-117, 1980. 

Dale, R. M. K., E. Martin, D. C. Livingston, and 

D. C. Ward, Biochemistry, 14, 2447-2457, 

Davidson, E. H., B. R. Hough, W. H. Klein, and 

R. J. Britten, Cell, 4, 217-238, 1975. 
Duncan, C, R. A. Biro, P. V. Choudary, J. T. 

Elder, R. R. C. Wang, B. G. Forget, J. K. De 

Riel, and S. A. Weissman, Proc. Nat. Acad. 

Sci. USA, 76, 5095-5099, 1979. 
Givens, J. F., and R. L. Phillips, Chromosoma, 

57, 103-117, 1976. 
Ingle, J., and J. Sinclair, Nature, 235, 30-32, 

Jendrisak, J., J. Biol. Chem., 255, 8529-8533, 

Kaback, D. B., L. M. Angerer, and N. Davidson, 

Nucl. Acids Res., 6, 2499-2517, 1979. 
Kuroiwa, A., and S. Natori, Nucl. Acids Res., 7, 

751-764, 1979. 
Leber, B., and V. Hemleben, Nucl. Acids Res., 7, 

1263-1281, 1979. 
Mathis, D., P. Oudet, and P. Chambon, Progr. 

Nucl. Acids Res. Mol. Biol, 24, 1-55, 1980. 
Murray, M. G., and W. F. Thompson, Plant Mol. 

Biol, in press. 
Murray, M. G., R. E. Cuellar, and W. F. Thomp- 
son, Biochemistry, 17, 5781-5790, 1978. 
Murray, M. G., D. L. Peters, and W. F. Thomp- 
son, J. Mol. Evol, 17, 31-42, 1981. 
Simon, J. H., and W. M. Becker, Biochim. Bio- 

phys. Acta, 454, 154-171, 1976. 
Southern, E. M.,J. Mol. Biol., 98, 503-518, 1975. 
Stalder, J. A., A. Larsen, J. D. Engel, M. Dolan, 

M. Groudine, and H. Weintraub, Cell, 20, 451- 

460, 1980. 


Marylee Everett, Richard A. Jorgensen, and William F. Thompson 

The pea leaf cDNA clones described 
previously (Year Book 79, 116-119) 
have been used to characterize changes 
in RNA populations extracted from 
apices of pea seedlings during greening 
and leaf development. Of particular in- 
terest are two of the cloned cDNAs, 
pEA5 and pEA28, which hybridize to 
RNAs that show large increases in 
abundance when RNAs from etiolated 
buds and light-grown leaves are com- 
pared. We have studied the time course 
of these changes and the role of phyto- 
chrome in the induction. 

RNA was extracted using the method 
of Glisen et al. (1974) from six-day-old 
etiolated pea buds and from similar pea 

buds greened by exposure to continu- 
ous white fluorescent light for periods 
ranging from 5 to 96 h. In addition, 
other batches of seedlings were exposed 
once a day during days 3, 4, and 5 of 
growth in darkness either to 3 min of 
red light or to 3 min of red light fol- 
lowed by 8 min of far-red light. On the 
sixth day, these plants were brought in- 
to white light, and RNA was extracted 
from their apices during greening. As 
an internal standard, a known amount 
of purified rabbit globin mRNA was 
added to these RNAs. They were then 
labeled with 32 P in vitro by the poly- 
nucleotide kinase reaction and hybrid- 
ized to recombinant plasmids contain- 



ing the different cDNAs, including 
globin cDNA, immobilized in bands 
on nitrocellulose filters, as described 
previously (Year Book 79, 116-119). 
Autoradiographs of these filters were 
scanned to yield quantitative data. 
Further information was obtained by 
hybridizing 32 P plasmid DNA to de- 
natured RNA, which was separated by 
electrophoresis through agarose gels 
and then blotted onto nitrocellulose by 
the method of Thomas (1980). 

RNAs homologous to pEA5 and 
pEA28 appear to be messages that are 
translated in vivo, since they are pres- 
ent on polysomes extracted from pea 
leaves. They are probably poly-adeny- 
lated because they are labeled by oligo- 
dT-primed reverse transcriptase in the 
presence of 8 mM pyrophosphate. 
Under these conditions, ribosomal 
RNA and chloroplast mRNA are la- 
beled at least 500 times less strongly 
than they are with kinase. On blots of 
denatured RNA, pEA5 hybridizes to a 
single band with a size of about 700 
nucleotides, and pEA28 hybridizes to a 
band of about 900 nucleotides and less 
strongly to a band of about 4400 nu- 

RNAs homologous to pEA5 and 
pEA28 are not detectable in etiolated 
buds above a background level of .001 % 
of the total RNA population. They are 
slightly above this background level in 
RNA extracted from tissue greened for 
20 h; by 96 h of growth in light they 
each represent about .02% of the total 
RNA. However, these amounts do not 
represent full induction, since in light- 
grown leaves they comprise .08% and 
.04% of total RNA, respectively. This 
slow development is also paralleled by 
increases in leaf fresh weight and total 
chlorophyll content, which do not reach 
the light-grown control values until 
after 1 20 h of growth in light. 

Since phytochrome promotes leaf de- 
velopment (Parker et al, 1949) and 
speeds chlorophyll synthesis in white 

light (Raven and Shropshire, 1975), the 
effect of phytochrome on the levels of 
these two transcripts was determined. 
Buds of the red-light-treated seedlings 
do not possess detectable levels of 
either of the RNAs, but by 24 h of 
greening in white light each RNA is 
present at the level found in non-pre- 
treated seedlings greened for 72 h. Far- 
red light after each red light exposure 
reverses this effect. As such reversibil- 
ity is characteristic of phytochrome ac- 
tion, the abundance of these two RNA 
transcripts is influenced either directly 
or indirectly by phytochrome action. 

The two transcripts described here 
are relatively abundant in mature leaf 
tissue though their function is un- 
known. They could be involved in the 
photosynthetic process, since their 
abundance follows the same pattern as 
the accumulation of chlorophyll. We be- 
lieve that other patterns of transcrip- 
tion in developing leaves will emerge 
when a recently constructed cDNA 
clone bank consisting of several hun- 
dred clones derived from etiolated, red- 
light-treated, and immature pea leaf 
RNA is characterized. Furthermore, in- 
vestigation of a similar clone bank from 
mung bean leaves will provide an inter- 
esting comparison because mung bean 
leaves are better developed in the dark 
than pea leaves and green significantly 
faster. The methods described above 
will be used to investigate these new 
clone banks and to document further 
examples of light-induced alterations 
in transcription. 


Glisen, V., R. Crkvenjakov, and C. Byus, Bio- 

chem., 13, 2633-2637, 1974. 
Parker, M., S. B. Hendricks, H. A. Borthwick, 

and F. Went, Amer. J. Bot, 36, 194-204, 1949. 
Raven, C„ and W. Shropshire, Photochem. and 

Photobiol, 21, 423-429, 1975. 
Thomas, P., Proc. Nat. Acad. Sci. USA, 77, 5201- 

5205, 1980. 






R. E. Cuellar and W. F. Thompson 

Last year (Year Book 79, 119-120), 
we reported preliminary observations 
of a set of cloned DNA fragments con- 
taining repetitive sequences from the 
pea genome. Among these clones were 
three, designated pPs 18, 42, and 84, 
which hybridized strongly to many of 
the same bands on Southern blots of 
pea DNA restriction fragments. This 
observation provided an initial indica- 
tion that these three independent 
clones might contain representatives of 
the same repetitive sequence family. 
Further analysis in the past year has 
indicated that the different cloned frag- 
ments do indeed contain common re- 
petitive elements. However, the clones 
are not homologous over their entire 
length, and the majority of sequences 
in each are quite distinct. Thus the 
cloned fragments do not represent iden- 
tical elements of a simple repeating 
family, but instead are examples of re- 
petitive sequences interspersed among 
other, unrelated, sequences in different 
regions of the genome. 

Our first indication that the three 
cloned fragments contained different 
as well as similar sequences came from 
the Southern blot analyses mentioned 
above, since in addition to seeing bands 
hybridizing with all three probes, we 
saw bands which hybridized specifically 
to only one of the clones. Differences in 
the sizes of the cloned pieces (8.2, 2.7, 
and 2.2 Kb for pPs 18, 42, and 84, re- 
spectively) were also consistent with a 
different sequence composition. When 
detailed restriction maps were con- 
structed for each clone, we could not dis- 
cern any regions of homology from the 
spacing of restriction sites for seven dif- 
ferent enzymes, leading us to conclude 
that the actual regions of interclone 
homology must be relatively short. 

Confirmation of this hypothesis 

came from direct cross-hybridization 
experiments in which restriction frag- 
ments from one clone were hybridized 
to filter-bound fragments from a differ- 
ent clone. In this way, we could locate 
regions of interclone homology on the 
physical map of the cloned DNA. It is 
clear that these regions are mostly 
quite short; estimates range between 
300 and 500 nucleotides. 

It seems necessary to postulate some 
type of chromosomal rearrangement 
process to account for the presence of 
many short homologous sequences in a 
wide variety of different sequence en- 
vironments. The importance of rear- 
rangements is further suggested by the 
observation that the linear order and 
spacing of homologous regions is clearly 
different in at least two of our three 

Each of the three cloned fragments 
also contains a number of sequences 
which are repeated to various extents 
in the pea genome but which are not 
common to the cloned DNAs. These se- 
quences have been characterized by 
various blot hybridization experiments 
involving the cloned DNA and total 
pea DNA. By hybridizing labeled pea 
DNA to filter-bound, restricted clone 
DNA, we can determine which regions 
of the clone contain sequences repeated 
in the pea genome. From the intensities 
of hybridization to different fragments, 
we can distinguish different repetition 

Each of these three clones contains 
much more repetitive DNA than is 
present in the common elements alone, 
and, in fact, there may not be any truly 
single-copy sequences represented. 
However, the repetition frequency of 
different elements within each clone 
does vary over a wide range, indicating 
that repetitive sequences of high and 



low copy number are closely inter- 
spersed in these regions. We have pre- 
viously concluded that extensive inter- 
spersion of short repeats with other 
repeats is the characteristic pattern of 
organization throughout more than 
90% of the entire pea genome (Murray 
etal, 1978, 1981). 

It is noteworthy that in spite of the 
interspersed pattern of organization 
evident in the clones, each fragment 
hybridizes predominantly to a set of 
discrete bands on Southern blots of 
restricted pea DNA. Partial digestion 
experiments show that many of these 
bands result from tandemly repeated ar- 
rays. Thermal dissociation profiles 
constructed by "melting" the hybrids 
off the pea DNA blots indicate that the 
hybrids formed in the band regions dis- 
play a range of thermal stabilities but 
are generally less stable than hybrids 
formed in self-reactions with filter- 
bound clone DNA. Together with our 
previous analysis of the clones them- 
selves, this evidence indicates that the 
tandemly repeated DNA sequences are 
related but not identical to the repeats 
in the clones. In other words, the clones 
we have studied contain represen- 
tatives of one or more repeat families 
containing both interspersed and tan- 
demly repeated members. 

Precisely this situation was envi- 

sioned on the basis of previous studies 
indicating considerable heterogeneity 
in pairing precision within repetitive 
sequence families of the pea genome 
(Preisler and Thompson, 1981; Year 
Book 79, 114-116). Thus, our studies 
with cloned DNA confirm in detail for a 
small sample of sequences a pattern of 
organization deduced from studies of 
the entire DNA complement. Both sets 
of data are consistent with a view of 
plant genome evolution as largely a 
stochastic process of amplification, di- 
vergence, rearrangement, and second- 
ary amplification, a process recently 
elaborated by ourselves (Murray et al, 
1981) and others (Flavell, 1980). It is 
noteworthy that the use of both cloning 
and whole-genome approaches lends a 
combination of scope and precision to 
those conclusions that would be impos- 
sible to achieve with either approach 


Flavell, R. B., Annu. Rev. Plant Physiol., 31, 
569-596, 1980. 

Murray, M. G., R. E. Cuellar, and W. F. Thomp- 
son, Biochemis try, 17, 5781-5790, 1978. 

Murray, M. G., D. L. Peters, and W. F. Thomp- 
son, J. Mol. Evol, 17, 31-42, 1981. 

Preisler, R. S., and W. F. Thompson, J. Mol. 
Evol, 17, 85-93, 1981. 


Jeffrey D. Palmer and William F. Thompson 

We have used clone banks of mung 
bean and pea chloroplast DNA in the 
construction of detailed restriction 
maps of these two genomes and in a 
comparison of the linear organization 
of homologous sequences in these and 
several other higher-plant chloroplast 
genomes. A detailed description of part 
of this research is in press (Palmer and 
Thompson, 1981a). 

A combination of double and triple 
digestions of individual cloned frag- 
ments of mung bean and pea chloro- 
plast DNA (Year Book 79, 120-123; 
Palmer and Thompson, 1981b), as well 
as hybridization of the cloned frag- 
ments to nitrocellulose filter-bound, 
total chloroplast DNA restriction frag- 
ments, has enabled us to map all the 
cleavage sites for the restriction en- 



donucleases BstEII, Kpn I, Pst I, 
Pvu II, Sac I, Sal I, Sma I, and Xho I 
on the circular chloroplast chromo- 
somes from mung bean and pea. In ad- 
dition, we have located the positions of 
the chloroplast ribosomal RNA genes 
on these maps by hybridizing 16S and 
23S chloroplast ribosomal RNA to 
filters containing chloroplast DNA 
restriction fragments. 

The mung bean chloroplast genome 
measures 150 kilobases in length; it in- 
cludes two identical sequences of 23 
kilobases which contain the ribosomal 
genes and which are arranged as an in- 
verted repeat separated by single-copy 
regions of 21 and 83 kilobases. The pea 
chloroplast genome is only 120 kilo- 
bases long, has only one set of ribo- 
somal genes, and does not possess any 
detectable repeated sequences. The 
mung bean inverted repeat structure is 
common to all other nonleguminous 
higher-plant chloroplast genomes so 
far studied, while the pea structure has 
been found only in the closely related 

legume— broad bean (Roller and De- 
lius, 1980). We conclude from these 
data that loss of one copy of the in- 
verted repeat sequence has occurred 
only rarely during the evolution of the 
Angiosperms and, in the case of the 
legumes, only after the divergence of 
the mung bean line from the pea- broad 
bean line. 

We have compared the order of ho- 
mologous sequences in the mung bean 
and pea chloroplast genomes by using 
each of 14 different cloned mung bean 
fragments, which together represent 
99% of the mung bean genome, as hy- 
bridization probes against filter-bound 
pea chloroplast DNA restriction frag- 
ments. A summary of these experi- 
ments (Fig. 41) indicates that a number 
of rearrangements (deletions, inser- 
tions, transpositions, inversions) have 
occurred which have scrambled the 
order of sequences common to the 
mung bean and pea chloroplast 
genomes. In similar sets of experi- 
ments we find that sequences shared 

16S 23S 

7.8 9.8 13.3 10.6 


16S 23S 


Fig. 41. Arrangement of homologous sequences in the mung bean and pea chloroplast genomes. 
Fourteen nonoverlapping, cloned mung bean restriction fragments representing 99% of the genome 
were each hybridized to replica nitrocellulose filters containing both pea and mung bean Pst I, Sal I, 
and Sma I fragments separated on a 0.7% agarose gel. The extent of the mung bean fragments used 
as probes is indicated by the two lines which converge above the fragments, while the size of the 
fragments in kb is given below. The pea fragments to which the mung bean probes hybridize are in- 
dicated by the lines leading from the mung bean fragments to the pea fragments. Wherever two dif- 
ferent mung bean fragments hybridized to the same pea fragment it was usually possible to differen- 
tiate which portion of the pea fragment hybridized to a given probe by determining which adjacent 
pea fragment(s) hybridized to the same probe. Restriction sites are represented by triangles (Pst I), 
squares (Sal I), and circles (Sma I). 



by the mung bean and broad bean 
chloroplast genomes, as well as by the 
pea and broad bean chloroplast ge- 
nomes, also show extensive rearrange- 

When we compare the organization of 
the mung bean chloroplast genome with 
those of several plants outside the leg- 
ume family, we find, quite surprisingly, 
that these plants share a strikingly 
similar pattern of chloroplast DNA se- 
quence organization. For example, we 
can detect only a single rearrangement 
(a large inversion of approximately 50 
kilobases of DNA within the large 
single-copy region) between mung bean 
and spinach chloroplast DNAs (Fig. 
42). The petunia and cucumber chloro- 
plast genomes are identical to that of 
spinach in gross organization, i.e., they 
also possess a large inversion within 
the single-copy region and in all other 
respects have the same organization as 
mung bean. The corn chloroplast 
genome also features the large inver- 
sion and, in addition, has one other re- 
arrangement—an apparent transposi- 
tion of several kilobases of DNA, again 
within the large single-copy region— 
that distinguishes it from the above 
four chloroplast genomes. 

At first glance, this situation may 
seem paradoxical: the linear order of 
shared sequences is far more conserved 
between mung bean and four species of 
plants outside the legume family than 
between mung bean and two other leg- 
umes—pea and broad bean. Note, how- 
ever, that mung bean, spinach (Herr- 
mann and Possingham, 1980), petunia 
(Bovenberg et al., 1981), cucumber (un- 
published data), and corn (Bedbrook 
and Bogorad, 1976), whose chloroplast 
genomes are very similar in sequence 
arrangement, all possess the inverted 
repeat sequence of 20-25 kilobases con- 
taining the ribosomal RNA genes. In 
contrast, extensive sequence rear- 
rangements are found only in com- 
parisons involving two plants, pea and 
broad bean, which have lost the in- 
verted repeat. In light of this correla- 
tion, we hypothesize that the inverted 
repeat is a strong stabilizing factor in 
chloroplast genome evolution, such 
that those chloroplast genomes pos- 
sessing the inverted repeat are very in- 
tolerant to gross sequence rearrange- 
ments, while those genomes which have 
lost the inverted repeat sustain re- 
arrangements at a much higher fre- 



Fig. 42. Arrangement of homologous sequences in the mung bean and spinach chloroplast 
genomes. Experimental conditions and figure nomenclature are the same as in Fig. 41, except that 
hybridization of mung bean clones was to spinach rather than pea Pst I, Sal I, and Sma I restriction 

department of plant biology 

Bedbrook, J. R., and L. Bogorad, Proc. Nat. 
Acad. Sci. USA, 73, 4309-4313, 1976. 

Bovenberg, W. A., A. J. Kool, and H. J. J. Nij- 
kamp, Nucl. Acids Res., 9, 503-517, 1981. 

Herrmann, R. G., and J. V. Possingham, in Re- 
sults and Problems in Cell Differentiation: the 


Chloroplast, J. Reinert, ed., pp. 45-96, 

Springer, Berlin, 1980. 
Roller, B., and H. Delius, Molec. Gen. Genet, 

178, 261-269, 1980. 
Palmer, J. D., and W. F. Thompson, Proc. Nat. 

Acad. Sci. USA, (in press), 1981a. 
Palmer, J. D., and W. F. Thompson, Gene, 15, 

21-26, 1981b. 



Holly L. Gorton and Winslow R. Briggs 

For phytochrome-mediated responses, 
tests of the Bunson-Roscoe reciprocity 
law have yielded basic information 
about the kind of photobiological con- 
trol involved. Classical low-fluence, 
reversible, phytochrome-mediated re- 
sponses in etiolated seedlings generally 
obey the reciprocity law. Light expo- 
sures of equal fluence elicit equal re- 
sponses whether they are delivered as 
short intense or long dim irradiations. 
The interpretation is that these low- 
fluence responses must be limited by a 
single photoproduct, presumably Pfr. 
High-irradiance phytocrome-mediated 
responses generally show reciprocity 
failure such that long dim exposures are 
more effective than short intense expo- 
sures of the same total fluence (Man- 
cinelli and Rabino, 1978). This pattern 
occurs because the light must be on for 
the high-irradiance responses to pro- 
ceed. When the light is switched off, the 
response stops. Together with other 
evidence, this reciprocity failure sug- 
gests a requirement for the presence of a 
low level of Pfr over a long time. 

We have tested the reciprocity law for 
stimulation of coleoptile and mesocotyl 
elongation and inhibition of antho- 
cyanin accumulation in the coleoptiles 
of light-grown corn in response to end- 
of-day, far-red (FR) irradiation. These 
responses are easily reversible and re- 
quire only low fluences, so one would ex- 
pect reciprocity to hold, as it does for 
the classical responses in etiolated 

seedlings. However, all three responses 
showed rapid and nearly complete 
reciprocity failure such that short in- 
tense irradiations were more effective 
than long dim ones of the same fluence 
{Year Book 79, 128-131). This is op- 
posite from the direction of reciprocity 
failure for the high-irradiance responses 
and hence is termed"reverse reciprocity 
failure." Red (R) reversal of these three 
end-of-day effects in corn also showed 
rapid reciprocity failure, and again, 
short intense irradiations were more ef- 
fective than long dim ones of the same 

The rapid, reverse reciprocity failure 
for FR and R is difficult to explain in 
terms of Pfr action and the current 
phytochrome dogma. Thus, it was of in- 
terest to determine whether rapid, 
reverse reciprocity failure is a common 
occurrence for responses to end-of-day 
irradiations in light-grown plants or 
whether it is an idiosyncracy restricted 
to corn. We have examined the reciproc- 
ity relationships of the end-of-day FR 
effects on oat coleoptile elongation, sun- 
flower and mung bean hypocotyl elon- 
gation, and anthocyanin synthesis in 
mung bean hypocotyls. We have also 
examined one additional response to 
end-of-day FR in corn— phytochrome- 
mediated stimulation of its own ac- 
cumulation. We have tested reciprocity 
for R reversal of the FR effects for all 
these responses except for hypocotyl 
elongation in sunflower. 








Q <" x -ft a 











O 3 e s s 

® 3 

S« c 

„ be 
X ffi o 

£ S 


(0 ft'S 

o o a as 
Q ^ 1 -a * 


8* E 
8 'So 

CO cu 


2 § 

ft -ri 

Co CO 

O "43 c « « 

-s « Eoa E 
co ^ n^ .a 


c 3 - 

J3 co O 
Eh t/J 




s * 

,-H ,-H ,-H o a> 













fH r-H rH !N 

Id <m io io 

o © © d 


lO lO lO 
N N N 

S 2 £ 2 
§ § § % 2 3 

'J3 'J3 >> 3 -3 3 

« « y 

be bD 2 

o o 

+ + 

O o o o 

3 -g, 3 



a > a 

oo g 

E ,-, * 
o o cp 

3 ■*■■ 
CO ro 



0Q O 

3; O 


_U o 

O i-l 


:- c 


£ -3 « 
cu E 3 

5 <M 3 

8 E 


B co & 

be 3 £ 

E E o 

>>3 u 


T3 2 7 ' 

c y co 

3 |,t 

-§-c £ 
° a u 

co e _ 

« £ 2 
£ ° £ 
3 ?. a 

CO 03 "^ CO 

2 1 <l> co 

a a co -u 

co 3 c xj 

£ ° 2 o 

i| & c 

H S « II 

CD n _ . 

H co C I 

«G C 3 co 


1 2 S S 

■u '3 cu O 

2 » I 8 

co o 

ST3 I 

S !3 C m 
CO O o >-i 

«-i a> 2 a> 

m — * ■*. -Z 

"O ^ ^ "c3 
c3 o SK—' 


"O co 

§ c 

25 i 

" J <h S" e 

£ S it R 

o m u 



The general experimental plan has 
been described (Gorton and Briggs, 
1980). To allow phytochrome measure- 
ment in the light-grown corn plants, 
seeds were imbibed and seedlings grown 
in the presence of the chlorosis-inducing 
herbicide Sandoz 9789 (0.2 mM) (Gorton 
and Briggs, 1980). For all other mea- 
surements, the seeds were imbibed and 
grown with water. Details of the growth 
and irradiation protocols varied for the 
different species, but all received end-of- 
day light treatments on at least three 
consecutive days. Oat coleoptiles and 
hypocotyls for mung bean and sun- 
flower seedlings were harvested for 
length measurement after 16 h of dark- 
ness, following the final end-of-day light 
treatment. Mung bean hypocotyls were 
then lyophilized and extracted for an- 
thocyanin measurement (Gorton and 
Briggs, 1980). 

Corn leaves were harvested 42 h after 
the final end-of-day treatment for 
phytochrome measurement. One-gram 
leaf samples were frozen in liquid nitro- 
gen and kept frozen up to one week. 
They were then ground to a fine powder 
and mixed with 400 mg CaC0 3 and 400 
fA buffer (1 M tris HC1, pH 7.5 with 300 
mM sodium ascorbate, 15 mM dithio- 
threitol, and 2.8 mM sodium diethyldi- 
thiocambamate). The final sample was a 
thick paste. Phytochrome was mea- 
sured as A (AA) between 660 and 730 
nm. There was no loss of spectro- 
photometrically detectable phyto- 
chrome during the storage period. Possi- 
ble scattering differences between 
samples were eliminated by the grind- 
ing and by the addition of CaC0 3 . 

Fluence-response curves for FR in- 
duction of these responses and for R 
reversal of a saturating FR exposure 
were obtained with a constant fluence 
rate for each experiment and varied 
duration of irradiation. Threshold and 
saturation values for FR and R fluence- 
response curves are summarized in 
Table 12 along with analogous data for 
the mesocotyl, coleoptile, and antho- 
cyanin responses in corn. Rapid, reverse 

reciprocity failure for the mesocotyl, co- 
leoptile, and anthocyanin responses in 
corn causes the fluence-response curves 
to be shifted to higher energy when low- 
fluence rates are used to obtain them, 
but a 25-fold difference in fluence rate is 
necessary to get a shift in the threshold 
of about one-half log unit (Gorton and 
Briggs, 1982). It is therefore unlikely 
that the small differences in fluence 
rates used would in themselves cause 
any significant shift in the fluence- 
response curves even if rapid, reverse 
reciprocity failure occurred. Also, since 
the responses are related to the log of 
the fluence, most of the differences in 
threshold and saturation between spe- 
cies and between responses are con- 
sidered minor. 

Reciprocity was tested for responses 
to end-of-day FR by varying fluence 
rate and duration of irradiation in- 
versely, but maintaining the same sub- 
saturating fluence. Reciprocity for R 
reversal was similarly tested for all 
responses except hypocotyl elongation 
in sunflower. Where reciprocity failed, 
it was necessary to determine if the 
failure could be explained by escape 
from photoreversibility (Year Book 79, 
128-131; Gorton and Briggs, 1982). 
For FR escape tests, plants were left in 
the dark for varying amounts of time 
between the end of the daily white-light 
period and the beginning of a short, 
high-fluence-rate FR irradiation of the 
same fluence used in the FR reciprocity 
experiments. Similar tests were done 
for escape from R reversibility by leav- 
ing the plants in the dark for varying 
amounts of time between saturating 
FR and the beginning of a short, high- 
fluence-rate R irradiation of the same 
fluence as that used in the R reciprocity 
tests. These results have been summar- 
ized along with the results from the 
corn coleoptile, mesocotyl, and antho- 
cyanin responses to the end-of-day FR 
(Year Book 79, 128-131; Gorton and 
Briggs, 1982) in Table 12. 

It is clear from Table 12 that rapid, 
reverse reciprocity failure not ex- 



plained by escape is not an ubiquitous 
phenomenon. For FR, reverse reciproc- 
ity failure occurred in all responses ex- 
amined in corn and oats, but for none of 
the responses tested in mung beans or 
sunflower. For R, reverse reciprocity 
failure was only observed for the meso- 
cotyl, coleoptile, and anthocyanin 
responses in corn {Year Book 79, 
128-131; Gorton and Briggs, 1982). 
The cause of the rapid, reverse reciproc- 
ity failure is unknown, but FR and R 
reverse reciprocity failure must stem 

from different causes, since they are 
separable in the cases of the coleoptile 
elongation response in oats and the 
phytochrome accumulation response in 


Gorton, H. L.. and W. R. Briggs, Plant Physiol., 

66, 1024-1026, 1980. 
Gorton, H. L., and W. R. Briggs, Plant, Cell and 

Environ., in press, 1982. 
Mancinelli, A., and J. Rabino, Bot. Rev., 44, 

129-180, 1978. 




Dina F. Mandoli and Winslow R. Briggs 

Two photoresponses in both the mes- 
ocotyl and coleoptile of Avena sativa 
(cv Lodi) have been quantitatively 
described (Mandoli and Briggs, 1981a). 
Although both the very-low- and the 
low-fluence response (VLF and LF, 
respectively) involve changes in elonga- 
tion rates of these organs in response to 
phytochome conversion, they can be 
distinguished on the basis of differ- 
ences in threshold and saturation 
energies of red light and percentage of 
Pfr required, far-red reversibility, and 
time at which red reciprocity fails. This 
report localizes the site(s) of photoper- 
ception for both the mesocotyl and co- 
leoptile when fluences for red light suf- 
ficient to saturate either VLF or LF 
responses are given to small areas of 
each organ. 

A masking technique was developed 
which allowed irradiation of just 2-mm 
regions of intact etiolated seedlings. 
Manipulations of the plants in dark- 
ness resulted in only about a 5% re- 
sponse in each tissue. Groups of plants 
for which different 2-mm regions were 
irradiated were allowed to grow in the 
dark for an additional 24 h until har- 
vest. The percentage response of meso- 
cotyl and coleoptile was graphed as a 

function of the position of the irradia- 
tion on the intact seedling. It was clear 
that a saturating illumination resulted 
in either a 100% VLF or LF response 
only if an area just below the node re- 
ceived light. 

If a discrete site(s) of photopercep- 
tion were present, we expected to ob- 
tain zero response when a nonsensitive 
region was irradiated and 100% re- 
sponse when a sensitive region was ir- 
radiated. Instead, we found that irradi- 
ation of any 2-mm region resulted in 
some response such that both the cole- 
optile and mesocotyl responses de- 
clined only gradually with distance 
from the most sensitive region for both 
VLF- and LF-saturating fluences. 
Either variation in length of the organs 
in the populations of seedlings irradi- 
ated in tissue light piping or scatter 
within the seedlings could reconcile 
this broad response pattern to discrete 
illumination with the idea of a discrete 
site of photoperception. However, since 
statistical analysis showed that the 
mesocotyl and coleoptile lengths were 
uniform in all seedlings used, this first 
explanation of the data was discarded. 

Red light was applied at 90° to the 
long axis of either mesocotyl or leaf tis- 



sue sections (coleoptile plus leaves). 
The amount of red light that emerged 
from the cut end of the tissue section 
was measured with a photomultiplier. 
All tissue sections were gently bent to 
prevent direct illumination of the pho- 
tomultiplier. The log of the axially 
transmitted light expressed as a func- 
tion of tissue length was the same for 
both mesocotyl and leaf tissues (y = 
-0.135* + 1.86, r 2 = 0.97). These 
organs can transmit laterally applied 
light over at least 25 mm, a distance 
greater than that expected if light were 
randomly scattered through the tis- 
sues. The presence of the coleoptilar 
node or tip attenuated the axially 
transmitted light relative to the 
amount transmitted through tissue 
sections devoid of these anatomical 
structures. Tissue damage also signifi- 
cantly decreased the amount of axially 
transmitted light. 

If we assumed there was a discrete 
site of photoperception, our quantita- 
tive data on light piping by etiolated 
tissues enabled us to localize more pre- 
cisely the site(s) of photoperception. 
First, we did a fluence-response curve at 
the 2-mm region, where 100% response 
for the VLF and LF was obtained. Sec- 
ond, we solved the light-piping equa- 
tion for tissue lengths (i.e., values of x) 
from to 19 mm distant from the hypo- 
thetical site of perception. This told us 
how much light was reaching the pho- 
tosensitive site, i.e., what the effective 
fluence was when we had irradiated dif- 
ferent 2-mm regions of the intact seed- 
lings. Third, we determined what the 
response of the seedling would have 
been had that fluence been given di- 
rectly to the 2-mm region that gave the 
maximum response (above), simply by 
reading the expected response from the 
fluence-response curve done with a 
2-mm irradiation field. The net result 
of these steps was a theoretical re- 
sponse pattern for 2-mm irradiations 
for which we knew precisely the loca- 
tion of the most sensitive site of percep- 
tion (where x = 0). This pattern could 

now be matched with our empirically 
obtained response pattern to 2-mm ir- 
radiation fields to determine the site of 
photoperception in the intact, etiolated 
seedlings. Light attenuation by coleop- 
tilar node and tip was accounted for in 
the final analysis. 

In conclusion, when the fiber optic 
properties of the seedling are known, 
the mesocotyl response can be ex- 
plained using one site of photopercep- 
tion within the mesocotyl itself. The co- 
leoptile, however, apparently depends 
on two discrete sites of photopercep- 
tion, one below the node (i.e., in meso- 
cotyl tissue) and one above the node 
(i.e., in coleoptile or leaf tissue). It is in- 
teresting to note that neither site of 
photoperception is correlated with the 
node or tip regions where phytochrome, 
as detected spectrophotometrically 
(Briggs and Siegelman, 1965; Pjon and 
Furuya, 1968) or immunologically 
(Pratt and Coleman, 1974), is most con- 
centrated. Hence, the total amount of 
phytochrome conversion is not the 
limiting parameter in the determina- 
tion of the photomorphogenic response. 
Although phytochrome is found 
throughout the etiolated oat seedling 
(Briggs and Siegelman, 1965; Pratt and 
Coleman, 1974), the state of most of the 
phytochrome is apparently immaterial 
to these two photomorphogenic re- 
sponses; only phytochrome conversion 
from about 4.5 to 6.5 mm below the 
node and from 1.5 to 2.5 mm above the 
node determines the photomorpho- 
genic response of the plants. 

The potential ecological significance 
of this light-piping property of etio- 
lated oats is also addressed (Mandoli 
and Briggs, 1981b). A biophysical 
study which deals directly with axial 
light transmission within these tissues 
is under way. 


Briggs, W. R., and H. W. Siegelman, Plant 
Physiol., 40, 934-941, 1965. 



Mandoli, D. F., and W. R. Briggs, Plant Physiol., Pjon, C. J., and M. Furuya, Planta, 81, 303-314, 

67, 733-739, 1981a. 1968. 

Mandoli, D. F., and W. R. Briggs, Plant, Cell and Pratt, L. H., and R. A. Coleman, Amer. J. Bot., 

Environ, in press, 1981b. 61, 195-202, 1974. 




Jonathan D. Walton, * James R. Shinkle, and Winslow R. Briggs 

During the study of red-light-induced 
changes in auxin binding in maize mes- 
ocotyls, it was discovered that Golgi- 
localized glucan synthetase activity 
(Ray et al, 1969), but not any of several 
other membrane marker activities, was 
lowered by red light irradiation of the 
seedling before extraction. We have 
done fluence-response curves and far- 
red reversibility experiments to study 
the relationship between the light- 
induced inhibition of growth (Vander- 
hoef and Briggs, 1979; Year Book 76, 
283-286) and the light-induced reduc- 
tion in glucan synthetase activity, and 
to test for phytochrome mediation of 
both responses. 

Unlike the response reported for the 
inhibition of mesocotyl growth in oat 
(Mandoli and Briggs, 1981), the maize 
variety used in our experiments ( WxWx 
X Bear Hybrid, CFS Research, 2761 
N. Main, Decatur, IL 62526) lacks the 
very-low-fluence response (VLFR) 
which is saturated at a fluence of 3 X 
10~ 3 nmol/cm 2 . A fluence of 10 -3 
nmol/cm 2 has no effect on growth (Fig. 
43). Other experiments indicate that 
the fluence-response curve in Fig. 43 is 
saturated at or near a red light fluence 
of 10 nmol/cm 2 . 

Golgi-localized glucan synthetase ac- 
tivity has been studied in several 
plants. The enzyme utilizes /xM 
concentrations of UDP-glucose as 
substrate, and requires Mg. In vitro, 
the enzyme produces j£M,4-linked 

♦Department of Biological Sciences, Stanford 
University, Stanford, CA 94305. 

glucan. In vivo, it probably makes 
hemicellulosic mixed-linkage glucan 
that is transported via the Golgi vesi- 
cles to the extracellular space. The 
activity of this enzyme in pea epicotyl 
tissue is controlled by auxin (Ray, 

The reduction of glucan synthetase 
activity by red light in the second cm 
below the node, but not in the apical 
cm, responds with the same threshold 

!B. a 

— r-^-i — ■ — i — ■ — 

1 1 ' 1 


- ^\i 









\ - 


^-*—//-l ' 1 ' 1 

I , I 

Loq 10 f 1 uence, nmo 1 /cm 

Fig. 43. Red light fluence-response curve for 
inhibition of growth of the maize mesocotyl. 
Plants were grown at 25 °C and 100% R. H. in 
complete darkness, irradiated at 72 h, and har- 
vested after an additional 12 h in darkness. No 
green safelights were used. Light sources were 
as described by Mandoli and Briggs (1981); max- 
imum fluence cate 1.0 nmol/cm 2 /s. Average 
mesocotyl length at time of irradiation was 
28.6 ±1.7 mm. Error bars represent ± SE, n = 



and saturation fluences as growth inhi- 
bition (Fig. 44; compare Fig. 43). The 
relative insensitivity of the apical cm of 
the mesocotyl may be related to the 
presence of meristematic cells. It is for- 
tuitous that the fluence-response 
curves for glucan synthetase and for 
growth have the same slope. Growth 
measured at the end of 12 h following 
irradiation (or dark control) is the inte- 
gral of growth rate (constant in dark- 
grown plants, changing in irradiated 
plants; see Vanderhoef and Briggs, 
1979; Year Book 76, 283-286), while 
glucan synthetase measurements rep- 
resent the level of activity only at the 
12-h time point. 

Far-red light alone significantly inhi- 
bits both growth and glucan synthe- 
tase (Table 13). Far-red light given 
after red light can reverse both effects 
of red light back to the level of far-red 
alone (Table 13). Green safelight causes 
partial inhibition both of growth and of 


1 — r^- 



1 ' 1 





• // 

• ■ ' ' ; 

0-1 cm 





■ ■ • 








\ 1-2 cm 










— Wa 

1 i 


1 , 1 


Loq 10 f 1 uence, nmol/cm 

Fig. 44. Red light fluence-response curve for 
reduction of Golgi-localized glucan synthetase 
activity in apical (0-1 cm below the node, dotted 
line) and sub-apical (1-2 cm below the node, solid 
line) segments of the mesocotyl. Irradiation pro- 
tocol is given in legend to Fig. 43. Average total 
particulate protein per apical segment was 68.8 
± 4.0 /xg/segment; per sub-apical segment 36.8 ± 
2.0 /xg/segment. Glucan synthetase was assayed 
as described by Ray et al., 1969. 

glucan synthetase level. A similar inhi- 
bition of mesocotyl growth by green 
light has been reported for oat (Mandoli 
and Briggs, 1981). In another experi- 
ment (not shown) a fluence of 126 
nmol/cm 2 of far-red light alone in- 
hibited growth and glucan synthetase 
by 43% and 39%, respectively. A 
comparison of these results with the 
data in Table 13 suggests that a far-red 
fluence of about 100 nmol/cm 2 is 
saturating. This result agrees well with 
the sensitivity of oat mesocotyl inhibi- 
tion by far-red light. 

In threshold fluence (Figs. 43 and 44) 
and timing (data not shown), the re- 
sponse of glucan synthetase activity 
corresponds closely with "first phase" 
inhibition of maize mesocotyl growth 
(Vanderhoef and Briggs, 1979; Year 
Book 76, 283-286). However, there are 
two discrepancies between the previ- 
ously obtained results (Vanderhoef et 
ai, 1979) and those reported here. 
First, the magnitude of the inhibition 
of mesocotyl growth by red light is 
higher in our results than in the pre- 
vious report. Second, the far-red-indu- 
cible mesocotyl growth inhibition is 
not seen by Vanderhoef et al. (1979). 
The fact that green safelights were em- 
ployed in the previous work could ac- 
count for these two discrepancies. 
Green-light-induced mesocotyl growth 
inhibition (Table 13) may totally ac- 
count for the increment of mesocotyl 
inhibition found here but not pre- 
viously reported. This green light 
fluence also may have saturated the 
far-red-inducible mesocotyl inhibition. 
In addition, the red-light-induced meso- 
cotyl inhibition reported by Vander- 
hoef et al. (1979) was measured after 24 
h, rather than after 12 h as reported 
here. Since in the present experiments 
growth rate inhibition by red light 
shows recovery within 24 h (data not 
shown), the apparent magnitude of the 
inhibition induced by red light could be 
significantly reduced by measuring 
growth after this period of time follow- 
ing irradiation. 



TABLE 13. Partial Reversibility by Far-Red Light of the Red-Light-Induced Inhibition of Growth 

and Glucan Synthetase* 



Glucan Synthetase 

% of Dark 


% of Dark 


(mm/12 h) 








































*The plants were grown in vermiculite for 78 h before irradiation and then returned to darkness 
for 12 h before harvest. Plants were handled under green safelight, which itself caused some inhibi- 
tion. Average mesocotyl length at time of irradiation was 35.8 ±1.3 mm. Light sources were as 
described by Gorton and Briggs (1980); total red light fluence = 9.0 nmol/cm 2 ; total far-red fluence 
= 846 nmol/cm 2 . Glucan synthetase was assayed in the subapical (1-2 cm below the node) cm of the 
mesocotyl. Average total particulate protein was 37.7 ±1.5 /xg/segment. 

Our results indicate that the effects 
of red light on inhibition of mesocotyl 
growth and on reduction of glucan syn- 
thetase activity are mediated at least 
partially by phytochrome (Table 13). In 
comparing the glucan synthetase inhi- 
bition and growth inhibition of the 
maize mesocotyl with the growth inhi- 
bition of the oat mesocotyl (Mandoli 
and Briggs, 1981), the threshold of the 
response is similar to the low-fluence 
response (LFR), but the far-red- and 
green-light-induced inhibition is char- 
acteristic of the very-low-fluence re- 
sponse (VLFR). Mandoli and Briggs 
(1981) explain the effect of far-red light 
alone by the absorption of far-red light 
by P r generating significant (1-3%) 
levels of P fr . Vanderhoef at al. (1979) 
reported a two-phase fluence-response 
curve, but their second phase is a high- 
irradiance response (HIR) which re- 
quires a high fluence and a long expo- 
sure time to potentiate the response. 

The physiological significance of the 
existence of multiple "phases" having 
different threshold requirements, and 
which are apparently induced by differ- 
ing levels of P fr , is unclear. These 
phases may be found to differ consider- 
ably between varieties, as does meso- 
cotyl growth (Inge and Loomis, 1937; 
Avery et al., 1937). The evidence from 

the experiments reported here suggests 
that such phases of mesocotyl inhibi- 
tion may be fused into a single, contin- 
uous fluence-response curve. This re- 
sult implies that, at least in the maize 
variety studied here, there is a contin- 
uous sensitivity of the mesocotyl inhi- 
bition to increasing levels of P fr , and 
that there may be a single mode of ac- 
tion of phytochrome in potentiating 
both the far-red reversible and far-red 
inducible portions of the fluence-re- 
sponse curve. 

The red light inhibition of mesocotyl 
growth is preventable by supplying 
auxin to irradiated shoots floated in so- 
lution (Vanderhoef and Briggs, 1978). 
Our experiments (data not shown) in- 
dicate that the effect of red light on 
glucan synthetase can be prevented by 
floating irradiated intact plants in solu- 
tion containing auxin. Thus, it appears 
that both glucan synthetase and 
growth are controlled, at least in part, 
by a phytochrome-modulated supply of 

Glucan synthetase activity in the 
mesocotyl at a particular time would be 
expected to correlate with growth rate 
and with levels of physiologically ac- 
tive auxin at that time. Therefore, the 
evidence presented here supports the 
suggestion that the activity of this en- 



zyme may be a useful indicator of the 
auxin status of a tissue. 


Avery, G. S., Jr., P. R. Burkholder, and H. B. 

Creighton, Bot. Gaz., 99, 25-143, 1937. 
Inge, F. D., and W. E. Loomis, Amer. J. Bot, 

24, 542-547, 1937. 

Mandoli, D. F., and W. R. Briggs, Plant Physiol., 

67, 733-739, 1981- 
Ray, P. M., Plant Physiol, 51, 601-608, 1973. 
Ray, P. M., T. L. Shininger, and M. M. Ray, Proc. 

Nat. Acad. Sci. USA, 64, 605-612, 1969. 
Vanderhoef, L. N., and W. R. Briggs, Plant 

Physiol., 61, 534-537, 1979. 
Vanderhoef, L. N., P. H. Quail, and W. R. Briggs, 

Plant Physiol, 63, 1062-1067, 1979. 
Walton, J. D., and P. M. Ray, Plant Physiol, 
1981, in press. 


Ta-Yan Leong and Winslow R. Briggs 

We have previously reported the pur- 
ification and solubilization of a blue- 
light-sensitive cytochrome-flavin com- 
plex from corn coleoptile membranes 
by differential centrifugation, sucrose 
density gradient centrifugation, and 
Renografin density gradient centrifu- 
gation (Year Book 79, 131-134; Year 
Book 79, 134-155; Leong and Briggs, 
1981). A specific 6-type cytochrome in 
the membrane preparation is reduced 
by blue light, and the reduction is prob- 
ably mediated by a flavin moiety 
(Briggs, 1980). 

The photoactive membrane fraction 
is shown to have Mg ++ -dependent 
ATPase activity which is not stimu- 
lated by K + and which is not inhibited 
by oligomycin, indicating that this 
membrane fraction is probably asso- 
ciated with the plasma membrane. 

The presence of flavin in the purified 
membrane fraction is indicated by the 
lumiflavin assay as well as by a fluores- 
cence emission maximum at 525 nm, 
obtained upon excitation at 450 nm. 

The redox potential of the Triton-sol- 
ubilized b cytochrome in the purified 
membrane fraction was measured in 
the presence of pyocyanine and potas- 
sium ferricyanide under anaerobic con- 
ditions. The redox states of both the 
pyocyanine and the b cytochrome were 
monitored simultaneously using a spec- 
trophotometer interfaced to a compu- 

ter. With increasing amount of dithion- 
ite, the pyocyanine peak between 600 
and 800 nm decreases in height, while 
the Soret difference peak of the b 
cytochrome at 427 nm increases in 
height. The ratios of log [ojc]/[red] as a 
function of the volume of dithionite 
added (which itself serves as a rough 
function of the potential) are then plot- 
ted for the pyocyanine and 6 cyto- 
chrome, as shown in Fig. 45. From the 
equation E h = E ml + RT/nF log 
[o;c]/[red], with —34 mV as the mid- 
point potential of pyocyanine at pH 7.0 
(Dutton and Wilson, 1974) and n = 2 
for pyocyanine, the midpoint potential 
of the b cytochrome is calculated to be 
-65 mV. 

The kinetics of the blue-light-induced 
reduction and dark reoxidation of the b 
cytochrome suggest that the midpoint 
potential of the b cytochrome is not af- 
fected by Triton X-100 solubilization, 
since the rates of these reactions were 
not significantly changed by Triton. 

From the above characteristics of the 
flavin-cytochrome complex, it is postu- 
lated that the complex is localized in 
the plasma membrane. When the flavin 
is excited by blue light, it acquires an 
electron (Massey, 1979), and on return 
to ground state, it passes the electron 
onto the 6 cytochrome. The midpoint 
potential of the b cytochrome will help 
to determine its position in this elec- 






"i — i — i — i — i — i — r 

~i — I — i — i — I — r 


J I I L 

20 30 40 50 

Volume of dithionite added, p. 

_L . J 1 i . 


Fig. 45. Titration of pyocyanine and b cytochrome against dithionite. The redox state of pyo- 
cyanine is estimated by the height of 600-800 nm peak; that of 6 cytochrome by the height of 
427-nm Soret peak. 

tron transport pathway as well as to 
provide a tool to study the mechanism 
of the photoreaction. 


Briggs, W. R., in Photoreceptors and Plant 
Development, 17-28, J. De Greef, ed., Ant- 
werpen University Press, 1980. 

Dutton, P. L., and D. F. Wilson, Biochim. 

BiophysActa, 346, 165-212, 1974. 
Leong, T.-Y., and W. R. Briggs, Plant Physiol, 

67, 1042-1046, 1981. 
Massey, V., in Photoreception and Sensory 

Transduction in Aneural Organisms, 253-269, 

F. Lenci and G. Colombetti, eds.. Plenum 

Press, New York, 1979. 





Ta-Yan Leong and Winslow R. Briggs 

Blue-light-induced phototropism in 
coleoptiles of grass seedlings has been 
extensively studied over the years (Thi- 
mann and Curry, 1960). Although a 
possible blue light photoreceptor sys- 
tem involving a flavin-cytochrome 
complex has been purified and charac- 
terized from corn coleoptiles (this 
Report), the physiological significance 
of this blue-light-sensitive membrane 
fraction with respect to phototropism 
is still unknown. 

Diphenyl ethers are herbicides 

known to inhibit electron transport at 
the plastoquinone-cytochrome f level in 
chloroplasts (Bugg et al, 1980), and it 
is known that light is required for them 
to be effective. Apparently, light en- 
ergy for activation is absorbed by xan- 
thophylls. Some toxic reaction(s) oc- 
curs that leads to a rapid increase in 
cell membrane permeability and subse- 
quent death of the tissues (Matsunaka, 

The effects of one of those diphenyl 
ethers, acifluorfen (Blazer), on the blue- 



light-induced absorbance change (re- 
duction of b cytochrome) in corn mem- 
brane preparations and on phototrop- 
ism and geotropism in etiolated oat 
seedlings, were investigated in this 

The presence of acifluorfen in the 
21,000 X g pellet from etiolated corn 
coleoptiles (see Goldsmith et al, 1980) 
leads to a stimulation in the amount of 
light-induced absorbance change in 
A(A 428 — A 410 ). Within the range of 
concentrations tested (from 1 X 10~ 6 
M to 1 X 10" 4 M), the stimulation in- 
creases log-linearly, as shown in Fig. 
46. The increase in magnitude of the 
light-induced absorbance change is 
caused by inhibition of the dark cyto- 
chrome rather than by any effect of the 
herbicide on the light reaction. 

The effect of acifluorfen on the blue- 
light-induced phototropic curvature in 
etiolated oat seedlings indicates that 
the phototropism response is sensitized 
by acifluorfen. In the presence of aci- 
fluorfen, the fluence-response curve 
(see Zimmerman and Briggs, 1963) is 
shifted in the direction of lower 
fluences by about 0.6 log unit. The 
degrees of curvature obtained in the 
control experiment at 1 X 10~ 5 and 1 X 
10 -4 M concentrations of acifluorfen 
after a 10 s exposure of 1 X 10 ~ 13 mol 
cm -2 s _1 blue light are compared in 
Table 14. Stimulations of 31% and 24% 
in the amount of curvature are obtained 
at 1 X 10" 5 and 1 X 10~ 4 M aci- 
fluorfen, respectively. 

Acifluorfen, at a concentration of 1 X 
10 -5 M, has no effect on the geotropic 
response in etiolated oat seedlings, in- 
dicating that acifluorfen is only active 

-6.2 -5.8 -5.4 -5.0 -4.6 -4.2 

Log [Acifluorfen, Ml 

Fig. 46. Effect of acifluorfen on the blue-light- 
induced absorbance change A (A 42 g ~ ^4io) in 
the 21KP from etiolated corn coleoptiles. 

TABLE 14. Effect on Acifluorfen on the Blue- 
Light-Induced Phototropism in Etiolated Oat 








1 X 10~ 5 

M Acifluorfen 



















Ave: 13.4 


1.31 ± 0.18 

1 X 10~ 4 

M Acifluorfen 













Ave: 12.4 


1.24 ± 0.21 

* Phototropism experiments on four-day -old 
oat seedlings were performed as described by 
Zimmerman and Briggs (1963). Acifluorfen, 
when present, was applied in the agar medium 
onto which 30-hour-old seedlings were 
transplanted. Data were from 10 s X 10 -13 mol 
cm -2 s _1 exposures. 



in the presence of light. The above re- 
sults suggest that the herbicide may be 
acting directly at the photoreceptor 

The good correlation between the ef- 
fect of acifluorfen at 1 X 10~ 5 M on the 
in vitro blue light reaction (Fig 46, 40% 
stimulation) and the in vivo phototro- 
pism response (Table 14, 31% stimula- 
tion) may indicate that the blue-light- 
induced absorbance change obtained 
with isolated membranes could be 
closely related to the phototropism ob- 
served in intact seedlings, and could 
thus be of physiological significance. 
The higher concentration (10 ~ 4 M) 

could be inhibiting other processes as 


Bugg, M. W., J. Whitmarsh, C. E. Black, and 
W. S. Cohen, Plant Physiol, 65, 47-50, 1980. 

Goldsmith, M. H. M., R. J. Caubergs, and W. R*. 
Briggs, Plant Physiol., 66, 1067-1073, 1980. 

Matsunaka, S., in Herbicides: Chemistry, 
Degradation and Mode of Action, 2nd Edition, 
Vol. 2, 709-739, Marcel Dekker, New York, 

Thimann, K. V., and G. S. Curry, in Comparative 
Biochemistry, 243-306, M. Florkin and H. S. 
Mason, eds., Academic Press, New York, 1960. 

Zimmerman, B. K., and W. R. Briggs, Plant Phy- 
siol, 38, 248-253, 1963. 




Siegrid Schoch, Holly L. Gorton, and Winslow R. Briggs 

When protochlorophyll(ide) is photo- 
transformed to chlorophyll(ide), the ini- 
tial product shows an absorption maxi- 
mum between 680 and 684 nm. During 
the next 10-40 min, this maximum 
shifts to about 670 nm. This spectral 
change was first described by (and 
named after) Shibata (1957). The pre- 
sent study addresses three questions 
concerning the Shibata shift. 

First, do carotenoids play a role 
either in determining the position of 
the absorption maximum immediately 
after phototransformation or on the 
Shibata shift itself? For this study (on 
corn), we used the herbicide Sandoz 
9789, a strong inhibitor of carotenoid 
synthesis in chloroplasts (B artels and 
McCullough, 1972) but one that has lit- 
tle effect on protochlorophyll(ide) for- 
mation or phototransformation (Jab- 
ben and Deitzer, 1978). Preliminary 
evidence suggesting that carotenoids 
could not play any major role in the 
Shibata shift appeared previously (Gor- 
ton and Briggs, Year Book 78, 

138-139). The current study presents 
more detailed analysis. 

Second, does the reduction of the ger- 
anylgeraniol ester of chlorophyll play 
any role in the shift? For this study (on 
oats), we used the technique of Schoch 
et al. (1980) for preventing reduction of 
the ester to phytol by anaerobiosis 
without significantly affecting the es- 
terification process itself. 

Finally, does proteolytic cleavage of 
the chlorophyll(ide) holochrome ac- 
count for the shift, as suggested by 
Stummann (1979)? For this study, we 
used the inhibitor of serine proteases, 
phenylmethylsulfonyl fluoride (PMSF) 
Fahrney and Gold, 1963), already 
known to be an effective inhibitor of 
proteolytic activity in oat extracts 
(Pike and Briggs, 1972). Since Butler 
arid Briggs (1966) had shown that the 
Shibata shift occurs, in glycerol-buffer 
homogenates of etiolated leaves, we 
used such preparations for the protease 
studies. The techniques for growing the 
seedlings, applying the inhibitors, and 



preparing the glycerol-buffer extracts 
are all described elsewhere, as are light 
sources and spectral methods (Schoch 
et at., 1981). 

The top two curves in Fig 47 il- 
lustrate room-temperature difference 
spectra obtained by subtracting the in 
vivo absorption spectrum for pro- 
tochlorophyll(ide) in etiolated corn 
leaves from the chlorophyll(ide) spec- 
trum obtained immediately following 
phototransformation. The two curves 
have not been normalized and both 
represent approximately equal 
amounts of tissue. It is clear that in- 
hibiting carotenoid synthesis to less 
than 1% of control amounts does not 
affect either the amount of protochloro- 

640 660 

Wavelength, nr 


Fig. 47. In vivo difference spectra from leaves 
from four-day-old etiolated corn (upper two 
curves) and from eight-day-old etiolated oats 
(lower two curves). Corn spectra from plants 
grown in the absence (above) or presence (below) 
of 0.2 mM Sandoz 9789. Oat spectra from plants 
kept in air (above) or under nitrogen (below) for 
16 h prior to harvesting. Samples measured at 
room temperature. Difference spectra obtained 
by subtracting spectrum obtained just before 
actinic irradiation from that obtained just after. 

phyll(ide) synthesized, the absorption 
spectrum of the protochlorophyll, or 
the absorption spectrum of the im- 
mediate product of phototransforma- 
tion. The lower difference spectra of 
Fig. 47 similarly illustrate that anaero- 
bic treatment known to prevent reduc- 
tion of the geranylgeraniol ester to 
phytol in oats likewise has no effect, 
either on protochlorophyll synthesis or 
on the spectral properties of both the 
protochlorophyll(ide) and the imme- 
diately formed chlorophyll(ide). 

Figure 48 illustrates room-tempera- 
ture difference spectra for the same two 
preparations shown in Fig. 47, but ob- 
tained by subtracting the chlorophyll 
(ide) curves obtained immediately after 
phototransformation from those ob- 
tained 20 min later, after the Shibata 
shift had taken place. It is clear that 
neither carotenoids nor geranlygera- 

Fig. 48. In vivo spectra as in Fig. 47, except 
obtained by subtracting spectrum obtained im- 
mediately after actinic irradiation from that 
taken after the Shibata shift had been permitted 
to occur for 20 min. 


TABLE 15. Relative Amounts of Four Major Spectral Components of Chlorophyll in Control and 
Nitrogen-Treated Plants Before and After the Shibata Shift 



Error f 662nm 




Control, 1 min L* 
Control, 1 min L, 10 min D* 
Nitrogen, 1 min L 
Nitrogen, 1 min L, 10 min D 



12 (ll) +t 

27 (12) 

40 (12) 

22 (13) 



14 (12) 


36 (12) 

13 (13) 




26 (12) 

38 (13) 

24 (13) 



17 (12) 

38 (12) 

32 (12) 

13 (13) 

* Abbreviations: D, dark; L, light; S.E., standard error. 

^Error: Factor by which difference between original curve and sum of components must be 
multiplied to give the error-difference curve produced by the RE SOL program. 
t+ Figures in parentheses denote half-bandwidths. 

niol esterification plays any role in the 
Shibata shift. 

For the study of a possible role of pro- 
tease activity, buffer-glycerol homoge- 
nates of oat leaves were used, and spec- 
tra were measured at liquid nitrogen 
temperature. PMSF at a concentration 
of 10~ 3 M, known to inhibit the activity 
of a neutral protease in oat extracts by 
about 80%, had no effect on the Shibata 
shift (data not shown). These experi- 
ments are not as conclusive as those 
with Sandoz 9789 or anaerobiosis, since 
it can be argued that the surviving 20% 
of protease activity that is insensitive 
to PMSF (Pike and Briggs, 1972) is 
that which is involved in the Shibata 

Spectra similar to the lower curves in 
Fig. 48 were obtained for oat glycerol- 
buffer extracts at liquid nitrogen tem- 
perature and subjected to curve analy- 
sis by the RESOL program (French et 
al, 1972). We wanted first to determine 
whether the Shibata shift in vitro could 
be modeled with the same four spectral 
components used by Virgin and French 
(1973) for their in vivo study of the 
Shibata shift and, second, if it could, 
whether this sensitive analytical 
technique could detect differences be- 
tween spectra obtained from control 
plants and from those plants subjected 
to anaerobiosis, differences undetect- 
able by straight difference spectro- 

scopy. The quantitative aspects of 
these analyses are shown in Table 15. It 
is clear that the Shibata shift can be 
largely accounted for by a loss of the 
684-nm component and its almost 
quantitative replacement by the 
669-nm component, as found by Virgin 
and French (1973) in their in vivo 
studies. It is also clear that the results 
are the same whether the plants were 
incubated in air or nitrogen prior to 
harvest. The results strongly support 
the conclusions that reduction of the 
geranylgeraniol ester of newly formed 
protochlorophyll(ide) has nothing to do 
with the Shibata shift. 


Bartels, P. G., and C. McCullough, Biochem. 

Biophys. Res. Commun., 48, 16-22, 1972. 
Butler, W. L., and W. R. Briggs, Biochem. 

Biophys. Acta, 112, 45-53, 1966. 
Fahrney, D. E., and A. M. Gold, J. Amer. Chem. 

Soc, 85, 997-1000, 1963. 
French, C. S., J. S. Brown, and M. C. Lawrence, 

Plant Physiol., 49, 421-429, 1972. 
Jabben, M., and G. F. Deitzer, Photochem. and 

Photobiol, 27, 799-802, 1978. 
Pike, C. S., and W. R. Briggs, Plant Physiol, 49, 

521-530, 1972. 
Schoch, S., C. Hehlein, and W. Rudiger, Plant 

Physiol., 66, 576-579, 1980. 
Shibata, K., J. Biochem. (Tokyo), 44, 147-173, 

Stummann, B. M., Physiol. Plant., 45, 122-126, 

Virgin, H. I., and C. S. French, Physiol. Plant., 

28, 350-357, 1973. 




Armond, P. A., see Badger, M. O.; Govind- 
jee; Malkin, S.; Zeiger, E. 

Avron, M., see Kaplan, A. 
679 Badger, M., O. Bjbrkman, and P. A. Ar- 
mond, An analysis of photosynthetic re- 
sponses and adaptation to temperature 
in higher plants: Temperature acclima- 
tion in the desert evergreen shrub Ner- 
ium oleander L., Plant, Cell and Envi- 
ron., in press. 

717 Belford, H. S., and W. F. Thompson, Sin- 

gle copy DNA homologies in Atriplex. 
I. Cross reactivity estimates and the role 
of deletion in genome evolution, Hered- 
ity, 46, 91-108, 1981. 

718 Belford, H. S., and W. F. Thompson, Sin- 

gle copy DNA in Atriplex. II. Hybrid 
thermal stabilities and molecular phy- 
logeny, Heredity, 46, 109-122, 1981. 

731 Berry, J. A., and W. J. S. Downton, Envi- 

ronmental regulation of photosynthesis, 
in Photosynthesis II. C0 2 Fixation 
and Plant Productivity, Govindjee, ed., 
Academic Press, in press. 

747 Berry, J. A., and J. K. Raison, Functional 
responses to temperature: Macrophytes, 
in Encyclopedia of Plant Physiology, 
New Series, Vol. 12 A, O. Lange, ed., in 
Berry, J. A., see Bloom, A. J.; Chapman, 
K. S. R.; Kaplan, A. 

760 Bjbrkman, O., Responses to different 
quantum flux densities, in Encyclopedia 
of Plant Physiology, New Series, Vol. 
12A, O. Lange, ed., 385-433, 1981. 

668 Bjorkman, O., Ecological adaptation of the 
photosynthetic apparatus, in Proc. 5th 
International Photosynthesis Congress, 
Greece, Sept. 6-13, in press, 1980. 
Bjorkman, O. see Badger, M. O.; Bloom, 
A. J.; Downton, W. J. S. 

712 Blatt, M. R., M. H. Weisenseel, and W. 
Haupt, A light-dependent current with 
chloroplast aggregation in the alga 
Vaucheria sessilis, Planta, in press. 

719 Bloom, A. J., H. A. Mooney, O. Bjbrkman, 

and J. A. Berry, Materials and methods 
for carbon dioxide and water exchange 
analysis, Plant, Cell and Environ., 3, 
371-376, 1980. 
733 Briggs, W. R., Photomorphogenesis, In 
McGraw-Hill Encyclopedia of Science 
and Technology, 5th Edition, S. P. 
Parker, ed., in press. 

732 Briggs, W. R., Phytochrome, In McGraw- 

Hill Encyclopedia of Science and Tech- 
nology, 5th edition, S. P. Parker, ed., in 
678 Briggs, W. R., Corn: Fuel for food or fuel 
for automobiles, BioScience, 31, 7, 1981. 
Briggs, W. R., see Caubergs, R.; Gorton, 

H. L.; Leong, T.Y.; Mandoli, D.; Quail, 
P. H. 

721 Brown, J. S., and S. Schoch, Spectral anal- 
ysis of chlorophyll-protein complexes 
from higher plant chloroplasts, Biochim. 
Biophys. Acta, in press. 

729 Caubergs, R., and W. R. Briggs, Blue light- 
induced absorbance changes in mem- 
brane preparations from corn coleop- 
tiles, in Photoreceptors and Plant Devel- 
opment, University of Antwerpen Press, 
J. De Greef, ed., pp. 67-78, 1980. 

726 Chapman, K. S. R., J. A. Berry, and M. D. 
Hatch, Photosynthetic metabolism in 
bundle sheath cells of the C 4 species Zea 
mays. Sources of ATP and NADPH and 
the contribution of photosystem II, 
Arch. Biochem. Biophys., 202, 330-341, 

761 Downton, W. J. S., O. Bjbrkman, and 
C. S. Pike, Consequences of increased 
atmospheric concentrations of carbon di- 
oxide for growth and photosynthesis of 
higher plants, in Carbon Dioxide and 
Climate: Australian Research, 1980, 
Australian Academy of Science, Can- 
berra, in press. 
Downton, W. J. S., see Berry, J. A.; Gov- 

704 Fork, D. C, and Govindjee, Chlorophyll a 
fluorescence transients of leaves from 
sun and shade plants, Naturwissen- 
schaften, 67, 510, 1981. 
Fork, D. C, see Govindjee; Malkin, S.; van 
Ginkel, G. 

751 Gorton, H. L., and W. R. Briggs, Phyto- 

chrome-mediated responses in light- 
grown corn showing rapid-reverse rec- 
iprocity failure, Plant, Cell and Environ., 
in press. 

752 Gorton, H. L., and W. R. Briggs, Rapid- 

reverse reciprocity failure for phyto- 
chrome control of its own synthesis, 
Plant, Cell and Environ., in press. 
720 Govindjee, W. J. S. Downton, D. C. Fork, 
and P. A. Armond, Chlorophyll a fluo- 
rescence transient as an indicator of 
water potential of leaves, Plant Sci. 
Lett, 20, 191-194, 1981. 
715 Govindjee, D. Fork, T. Wydryznski, M. 
Spector, and D. G. Winget, Photosystem 
II reactions in liposomes reconstituted 
with cholate-extracted thylakoids and a 
manganese-containing protein, Photo- 
biochem. and Photobiophys., 1, 347- 
351, 1980. 

Govindjee, see Fork, D. C. 

Harvey, G. W., see Melis, A. 

Hatch, M. D., see Chapman, D. S. R. 

Haupt, W., see Blatt, M. R. 
711 Kaplan, A., and J. A. Berry, Glycolate 



excretion and the 2 /C0 2 net exchange 
ratio during photosynthesis in Chlamy- 
domonas reinhardtiu Plant Physiol., 67, 
229-232. 1981. 
683 Kaplan, A., A. Schreiber, and M. Avron, 
Salt induced metabolic changes in Dun- 
aliella. Plant Physiol, 96, 185-188, 1981. 

713 Leong, T.-Y., and W. R. Briggs, Partial 

purification and characterization of a 
blue light-sensitive cytochrome-flavin 
complex from corn membranes, Plant 
Physiol, 67, 1042-1046. 
745 Leong, T.-Y., W. R. Briggs, and R. N. 
Vierstra, A blue light-sensitive cyto- 
chrome flavin complex from coleoptiles. 
Further characterization, Photochem. 
and Photobiol, in press. 

736 Levitt, J., Correspondence: Response to 

Stout's Answer, Plant, Cell and En- 
viron., 3, 291-292, 1980. 

669 Malkin, S., P. A. Armond, H. A. Mooney, 
and D. C. Fork, Photosystem II photo- 
synthetic unit sizes from fluorescence 
induction in leaves. Correlation to photo- 
synthetic capacity. Plant Physiol, 67, 
570-579, 1981. 

705 Malkin, S., and D. C. Fork, Photosynthetic 
units of sun and shade plants, Plant 
Physiol, 67, 580-583, 1981. 

707 Mandoli, D., and W. R. Briggs, Photo- 
chrome control of two low irradiance 
responses in etiolated oat seedlings. 
Plant Physiol, 67, 733-739, 1981. 

755 Melis, A., and G. W. Harvey, Regulation 
of photosystem stoichiometry, chloro- 
phyll a and chlorophyll b content and 
its relation to chloroplast ultrastructure, 
Biochim. Biophys. Acta, 637, 138-145, 
Melis, A., see Zeiger, E. 
Mooney, H. A., see Bloom, A. J.; Mal- 
kin, S. 

702 Murray, M. G., D. Peters, and W. L. 
Thompson, Ancient repeated sequences 
in pea and mung bean genomes and a 
model for genome evolution, J. Mol 
Evol, 17, 31-42, 1981. 

714 Murray, M. G., and W. F. Thompson, 

Rapid isolation of high molecular weight 
plant DNA, Nucleic Acids, 8, 4321-4325, 

737 Palmer, J. D., and W. F. Thompson, Clone 

banks of the mung bean: Pea and spin- 
ach chloroplast genomes. Gene, in press. 
738 Palmer, J. D., and W. F. Thompson, Re- 
arrangements in the chloroplast ge- 
nomes of mung bean and pea, Proc. Nat. 
Acad. Sci. USA, in press. 

Peters, D. L., see Murray, M. G. 

Pike, C. S., see Downton, W. J. S. 

708 Preisler, R. S., and W. F. Thompson, Evo- 

lutionary sequence divergence within re- 
peated DNA families of higher plant 
genomes. I. Analysis of reassociation 
kinetics, J. Mol. Evol, 17, 78-84, 1981. 

709 Preisler, R. S., and W. F. Thompson, Evo- 

lutionary sequence divergence within re- 
peated DNA families of higher plant 
genomes. II. Analysis of thermal de- 
naturation, J. Mol. Evol, 17, 85-93, 1981. 
730 Quail, P. H., and W. R. Briggs, Irradiation- 
enhanced phytochrome pelletability: Fil- 
tration and mixing experiments with 
Avena, in Photoreceptors and Plant 
Development, Antwerpen University 
Press, J. De Greef, ed., pp. 167-168, 

Raison, J. K., see Berry, J. A.; van Gin- 
kel, G. 

Schoch, S., see Brown, J. S. 

Schreiber, A., see Kaplan, A. 

Spector, M., see Govindjee. 

Thompson, W. F., see Belford, H. S.; Mur- 
ray, M. G.; Palmer, J. D.; Preisler, R. S. 
759 van Ginkel, G., and D. C. Fork, The effect 
of ageing in vitro of spinach chloroplasts 
on the phase transition temperatures 
of thylakoid membranes, Photobiochem. 
and Photobiophys., 2, 239-243, 1981. 
716 van Ginkel, G., and J. Raison, Light- 
induced formation of 2 oxygen radicals 
in systems containing chlorophyll, 
Photochem. and Photobiol, 32, 793-798, 

Vierstra, R. D., see Leong, T.-Y. 

Weisenseel, M. H., see Blatt, M. R. 

Winget, D. G., see Govindjee. 

Wydryznski, T., see Govindjee. 

710 Zeiger, E., P. A. Armond, and A. Melis, 

Fluorescence properties of guard cell 
chloroplasts: Evidence for linear electron 
transport and light-harvesting pigments 
of PSI and PSII, Plant Physiol, 67, 
17-20, 1981. 


Berry, Joseph A., Photosynthesis at high tem- 
perature: The role of the chloroplast mem- 
brane, Department of Pharmacology, Stan- 
ford University, Stanford, California, January 
22, 1981. 

Berry, Joseph A., Environmental control of 

photosynthesis, USDA Western Regional 
Laboratory, Albany, California, May 7, 1981. 
Berry, Joseph A., Control of photosynthetic 
rate by environmental and biochemical fac- 
tors, Department of Botany, The Hebrew 
University of Jerusalem, Israel, May 22, 1981. 



Bjorkman, Olle, Ecological adaptations of the 
photosynthetic apparatus, Fifth International 
Photosynthesis Congress, Athens, Greece, 
September 10, 1980. 

Bjorkman, Olle, Ecological aspects of photo- 
synthetic productivity. First Binational 
Japan-Australia Symposium, Nagoya, Japan, 
April 21-27, 1981. 

Briggs, Winslow R., A blue light-sensitive fla- 
vin-cytochrome complex from corn and Neuro- 
spora—a candidate for blue light photorecep- 
tor, Boyce Thompson Institute for Plant 
Research, Cornell University, Ithaca, New 
York, October 8, 1980. 

Briggs, Winslow R., Phytochrome regulation of 
elongation in corn and oats— some surprises, 
Division of Biological Sciences, Cornell Uni- 
versity, Ithaca, New York, October 10, 1980. 

Briggs, Winslow R., Phytochrome regulation of 
elongation in corn and oats— some surprises, 
Fall Meeting of the Washington, D.C., section 
of the American Society of Plant Physiolo- 
gists, Washington, D.C., October 15, 1980. 

Briggs, Winslow R., Spectral studies of a blue 
light-sensitive cytochrome-flavin complex 
from corn membranes. Shell Development Co., 
Modesto, California, January 9, 1981. 

Briggs, Winslow R., A light-sensitive cyto- 
chrome-flavin complex from corn membranes, 
Chemical Biodynamics, University of Cali- 
fornia, Berkeley, February 4, 1981. 

Briggs, Winslow R., A light-sensitive flavin- 
cytochrome complex from corn membranes: 
The blue light photoreceptor?, Monsanto 
Agricultural Products Co., St. Louis, 
Missouri, March 20, 1981. 

Briggs, Winslow R., A light-sensitive flavin- 
cytochrome complex from higher plants and 
fungi, Department of Microbiology and Cell 
Science, University of Florida, Gainesville, 
May 18, 1981. 

Briggs, Winslow R., What's new in plant biol- 
ogy?, Division of Biology, California Institute 
of Technology, Pasadena, May 26, 1981. 

Briggs, Winslow R., see Gorton, H. L.; Leong, 
T.Y.; Mandoli, D. F. 

Brown, Jeanette S., Stoichiometry of system I 
and II reaction centers and of plastoquinone 
in different photosynthetic membranes, Third 
International Conference on Photochemical 
Conversion & Storage of Solar Energy, Uni- 
versity of Colorado, Boulder, August 4, 1980. 

Brown, Jeanette S., Organization of chlorophyll 
in plants, CSIRO-Division of Fisheries and 
Oceanography, Cronulla, Australia, February 
27, 1981. 

Brown, Jeanette S., Absorption and fluores- 
cence of chlorophyll in flashed leaf chloro- 
plasts, Australian National University, 
Canberra City, March 18, 1981. 

Brown, Jeanette S., Spectroscopy of chlorophyll 
in vivo, Australian National University, 
Canberra City, March 24, 1981. 

Brown, Jeanette S., Computer technology 
series. Practical side of computer assisted 
analysis of spectral data, CSIRO-Division of 
Plant Industry, Canberra City, Australia, 
April 3, 1981. 

Brown, Jeanette S., Stoichiometry of system I 
and system II reaction centers and of plastoqui- 
none in different photosynthetic membranes, 
Institute of Applied Microbiology, University 
of Tokyo, Hongo, Japan, May 12, 1981. 

Brown, Jeanette S., Chlorophyll and the two 
light reactions of photosynthesis, Biology 
Department, Silliman University, Dumaguete 
City, Negros Or., Republic of the Philippines, 
April 28, 1981. 

Brown, Jeanette S., see Melis, A. 

Critchley, C, see Stephen B. Powles 

Cuellar, Richard E., Cloned restriction frag- 
ments of pea DNA containing arrays of repet- 
itive sequence elements, Gordon Conference, 
Andover, New Hampshire, June 11, 1980. 

Cuellar, Richard E., and Richard A. Jorgensen, 
Cloned fragments of pea and mung bean DNA 
containing repeated DNA sequences, 
American Society of Plant Physiologists, 
Pullman, Washington, August 6, 1980. 

Cuellar, Richard E., Molecular clones of higher 
plant DNA for the study of plant genome 
evolution, Stanford-Carnegie Plant Science 
Seminars, Carnegie Institution, Stanford, 
California, December 10, 1980. 

Fork, David C, The influence of the physical 
phase of thylakoid membrane lipids on photo- 
synthetic activity, Texas A&M University, 
College Station, November 11, 1980. 

Fork, David C, The effects of temperature- 
induced structural alterations on photosyn- 
thetic activity of thylakoid membranes, De- 
partment of Botany, University of Texas, 
Austin, November 13, 1980. 

Fork, David C, Temperature control of photo- 
synthesis, Biology Department, Stanford Uni- 
versity, Stanford, California, June 1, 1981. 

Gorton, Holly L., and Winslow R. Briggs, Phy- 
tochrome-mediated anthocyanin formation in 
coleoptiles of light-grown corn, American 
Society of Plant Physiologists, Pullman, 
Washington, August 4, 1980. 

Gorton, Holly L., Phytochrome-mediated re- 
sponses to end-of-day irradiations showing 
rapid, reverse reciprocity failure, Stanford- 
Carnegie Plant Science Seminars, Carnegie In- 
stitution, Stanford, California, April 22, 1981. 

Gorton, Holly L., and Winslow R. Briggs, Rapid 
reverse reciprocity failure for phytochrome 
control of its own synthesis, American Society 
of Plant Physiologists, Laval University, Ste. 
Foy, Quebec, Canada, June 14-18, 1981. 

Harvey, Geoffrey W., Photosynthetic perfor- 
mance of isolated leaf cells from sun and shade 
plants, American Society of Plant Physiolo- 
gists, Pullman, Washington, August 6, 1980. 

Jorgensen, Richard A., Time- and tissue-specific 



expression of individual genes in Pisum, 
American Society of Plant Physiologists, Pull- 
man, Washington, August 6, 1980. 

Jorgensen, Richard A., Applying molecular 
genetic approaches to problems in plant evo- 
lution and development, Department of Bac- 
teriology and Plant Growth Laboratory, 
University of California, Davis, October 31, 

Jorgensen, Richard A., .see Cuellar, Richard E. 

Katoh, Sakae, see Satoh, Kazuhiko. 

Roller, Dov, Solar tracking in leaves, Thimann 
Laboratory, University of California, Santa 
Cruz, October 30, 1980. 

Roller, Dov, Solar tracking in leaves, Depart- 
ment of Botany, University of Washington, 
Seattle, November 14, 1980. 

Roller, Dov, Solar tracking in leaves. Labora- 
tory of Chemical Biodynamics, University of 
California, Berkeley, January 21, 1981. 

Roller, Dov, Solar tracking by leaves, Depart- 
ment of Vegetable Crops and Graduate Divi- 
sion, Distinguished Visiting Speaker Lec- 
tures, University of California, Davis, 
February 11, 1981. 

Roller, Dov, Solar tracking by leaves, Stan- 
ford-Carnegie Plant Science Seminars, Car- 
negie Institution, Stanford, California, June 3, 

Leong, Ta-Yan, and Winslow R. Briggs, Puri- 
fication of a light-sensitive cytochrome-flavin 
complex from etiolated corn coleoptiles, 
American Society of Plant Physiologists, Pull- 
man, Washington, August 4, 1980. 

Leong, Ta-Yan, Winslow R. Briggs, and Richard 
D. Vierstra, Spectral studies on a blue light- 
sensitive cytochrome-flavin complex from 
corn membranes, European Symposium on 
Light-Mediated Plant Development, Bischofs- 
mais, West Germany, April 7, 1981. 

Leong, Ta-Yan, Richard D. Vierstra, and 
Winslow R. Briggs, Properties of a light- 
sensitive cytochrome-flavin complex from 
etiolated corn coleoptiles, American Society of 
Plant Physiologists, Laval University, Ste. 
Foy, Quebec, Canada. June 14-18, 1981. 

Levitt, Jacob, Problems associated with assess- 
ing crop losses due to water stress. University 
of Minnesota, Department of Plant Pathology 
(Crop Loss Assessment), E. C. Stakman Com- 
memorative Symposium, Minneapolis, Au- 
gust 22, 1980. ' 

Levitt, Jacob, Membrane function in plant 
responses to environmental stress— the role of 
proteins and lipids. University of Missouri, 
Columbia, October 23, 1980. 

Mandoli, Dina F., and Winslow R. Briggs, 
Photomorphogenesis of etiolated Auena 
sativa (cv Lodi), American Society of Plant 
Physiologists, Pullman, Washington, August 
4, 1980. 

Mandoli, Dina F., and Winslow R. Briggs, 
Etiolated oat tissues act as conduits for light, 
American Society of Plant Physiologists, 

Laval University, Ste. Foy, Quebec, Canada, 
June 14-18, 1981. 

Mandoli, Dina F., and Winslow R. Briggs, Pho- 
toreceptive regions of the low-irradiance 
responses in etiolated oats, American Society 
of Plant Physiologists, Laval University, Ste. 
Foy, Quebec, Canada, June 14-18, 1981. 

Melis, Anastasios, Biphasic reduction kinetics 
of the primary electroacceptor: quinone of 
photosystem II. Evidence for two distinct 
photochemical centers in system II of chloro- 
plasts, Quinone Meeting at Dartmouth Med- 
ical School, Hanover, New Hampshire, Oc- 
tober 20, 1980. 

Melis, Anastasios, and Jeanette S. Brown, Spec- 
trophotometric determination of the stoichio- 
metry of reaction centers in different 
photosynthetic membranes, American Society 
of Plant Physiologists, Pullman, Washington, 
August 4, 1980. 

Murray, Michael G., The distinction between 
coding and single-copy sequences in the pea 
genome, American Society of Plant Physiolo- 
gists, Pullman, Washington, August 6, 1980. 

Murray, Michael G., DNA sequence organiza- 
tion in coding and uncoding regions of the pea 
genome, International Plant Research In- 
stitute, San Carlos, California, February 5, 

Murray, Michael G., DNA sequence organiza- 
tion in coding and uncoding regions of the pea 
genome, Botany Department, University of 
Wisconsin, and Agrigenetics Corporation, 
Madison, March 3, 1981. 

Murray, Michael G., DNA sequence organiza- 
tion in coding and uncoding regions of the pea 
genome, Monsanto Corporation, St. Louis, 
Missouri, April 4, 1981. 

Murray, Michael, G., Pea ribosomal chromatin 
structure, Gordon Conference, Plant Molecu- 
lar Biology, Andover, New Hampshire, June 
8, 1981. 

Palmer, Jeffrey D., and William F. Thompson, 
Chloroplast DNA levels in higher plants, 
American Society of Plant Physiologists, Pull- 
man, Washington, August 7, 1980. 

Palmer, Jeffrey D., and William F. Thompson, 
Construction of complete clone banks of the 
mung bean, pea, and spinach chloroplast 
genomes and their use in studying chloroplast 
DNA evolution, American Society of Plant 
Physiologists, Pullman, Washington, August 
7, 1980. 

Palmer, Jeffrey D., Evolution of higher plant 
chloroplast and mitochondrial DNA, Rocke- 
feller University, New York, New York, 
December 29, 1980. 

Palmer, Jeffrey D., Evolution of higher plant 
chloroplast and mitochondrial DNA, Stan- 
ford-Carnegie Plant Science Seminars, Carne- 
gie Institution, Stanford, California, February 
4, 1981. 

Palmer, Jeffrey D., Evolution and organization 
of the higher plant chloroplast and mitochon- 



drial genomes, First Annual Congress on 
Recombinant DNA, San Francisco, California, 
February 28, 1981. 

Powles, Stephen B., and Christa Critchley, 
Photoinhibition in beans adapted at high and 
low light regimes, American Society of Plant 
Physiologists, Pullman, Washington, August 
5, 1980. 

Powles, Stephen B., Photoinhibition: The de- 
structive effect of light on photosynthesis, 
Department of Botany, University of Califor- 
nia, Davis, October 30, 1980. 

Powles, Stephen B., Interaction between light 
and chilling temperatures: Photooxidation 
and photoinhibition, Stanford-Carnegie Plant 
Science Seminars, Carnegie Institution, Stan- 
ford, California, November 19, 1980. 

Powles, Stephen B., Photoinhibition in intact 
leaves of C 4 plants, Fifth International Photo- 
synthesis Congress, Athens, Greece, Septem- 
ber 9, 1980. 

Powles, Stephen B., The photoinhibition phe- 
nomenon in higher plants, University of 
Kaiserslautern, West Germany, September 
12, 1980. 

Powles, Stephen B., The photoinhibition phe- 
nomenon in higher plants, Dow Chemical 
Company, Wayland, Massachusetts, June 10, 

Powles, Stephen B., Temperature and light 
stress in higher plants, Pfizer Central 
Research, Groton, Connecticut, June 13, 1981. 

Preisler, Richard S., and William F. Thompson, 
Evolutionary sequence divergence within 
repeated DNA families of peas and mung 
beans, American Society of Plant Phys- 
iologists, Pullman, Washington, August 6, 

Reichert, Terri L., Specificity of several auxin- 
mediated responses: Implications for auxin 
action, Stanford-Carnegie Plant Science 
Seminars, Carnegie Institution, Stanford, Cal- 
ifornia, February 18, 1981. 

Satoh, Kazuhiko, Induction phenomena related 
to the photosynthetic electron transport 
system, Stanford-Carnegie Plant Science 
Seminars, Carnegie Institution, Stanford, 
California, January 28, 1981. 

Satoh, Kazuhiko, and Sakae Katoh, The cause 
of the DPSj change of chlorophyll fluores- 
cence, American Society of Plant Physi- 
ologists, Laval University, Ste. Foy, Quebec, 
Canada, June 15, 1981. 

Schoch, Siegrid, Investigations of chlorophyll 
synthesis and degradation in higher plants 
and Euglena, Stanford-Carnegie Plant Science 
Seminars, Carnegie Institution, Stanford, 
California, November 5, 1980. 

Seemann, Jeffrey R., Winter annuals in Death 
Valley, Stanford-Carnegie Plant Science 
Seminars, Carnegie Institution, Stanford, 
California, March 4, 1981. 

Seemann, Jeffrey R., High photosynthetic 
capacities of desert winter annuals measured 
in situ and the relationship to protein content 
and RuBP carboxylase kinetic properties, 
American Society of Plant Physiologists, 
Laval University, Ste. Foy, Quebec, Canada, 
June 15, 1981. 

Smith, Celia M., Aspects of the physiological 
ecology of Porphyra perforata, a red mid- 
intertidal alga, American Society of Plant 
Physiologists, Laval University, Ste. Foy, 
Quebec, Canada, June 16, 1981. 

Thompson, William F., Applications of recom- 
binant DNA technology to problems in plant 
molecular genetics, Society of Sigma Xi, 
Modesto, California, January 21, 1981. 

Thompson, William F., Comparative aspects of 
genome structure in higher plants, Shell 
Development Company, Modesto, California, 
January 22, 1981. 

Thompson, William F., Structure and function 
of DNA in higher plants, Biosystematists 
Society, Stanford University, Stanford, Cali- 
fornia, February 10, 1981. 

Thompson, William F., see Richard Cuellar; 
Richard Jorgensen; Jeffrey D. Palmer; Michael 
G. Murray; Richard S. Preisler. 

Vierstra, Richard D., see Ta-Yan Leong. 

Williams, William E., Optimal water-use effi- 
ciency in a California shrub, American Society 
of Plant Physiologists, Laval University, Ste. 
Foy, Quebec, Canada, June 14-18, 1981. 




Research Staff 
Joseph A. Berry 
Olle Bjorkman 
Winslow R. Briggs, Director 
Jeanette S. Brown 
David C. Fork 

C. Stacy French, Director Emeritus 
William M. Hiesey, Emeritus 
Malcolm A. Nobs 
William F. Thompson 

Research Associate 
Michael G. Murray 

Postdoctoral Fellows 

Jacob Levitt, Senior Fellow, University 
of Minnesota, Minneapolis 

Stephen B. Powles, CSIRO, Canberra, 
Australia 1 

Siegrid Schoch, Senior Fellow, Univer- 
sity of Munich, W. Germany 2 

Carnegie Institution of Washington 
Postdoctoral Fellows 

Marylee Everett, Senior Fellow, Merri- 
mack College, MA 

Geoffrey W. Harvey 3 

George E. Hoch, Senior Fellow, Univer- 
sity of Rochester, NY 4 

Richard A. Jorgensen 

Dov Roller, Senior Fellow, The Hebrew 
University of Jerusalem, Israel 

Ta-Yan Leong 

Anastasios Melis 5 

Gunnar Oquist, Senior Fellow, Univer- 
sity of Umea, Sweden 6 

Stephen B. Powles 1 

Kazuhiko Satoh, Senior Fellow, Univer- 
sity of Tokyo, Japan 

Siegrid Schoch, Senior Fellow, Univer- 
sity of Munich, W. Germany 2 

'CSIRO Fellow to December 31, 1980; Carnegie 
Institution of Washington Fellow thereafter. 

2 Deutsche Forschungsgemeinschaft Fellow to 
October 31, 1980; Carnegie Institution of Wash- 
ington Fellow to February 28, 1981. 

3 To November 30, 1980 

4 To August 31, 1980 

5 To March 31, 1981 

6 To December 31, 1980 

Graduate Students 

Richard E. Cuellar, Stanford Univer- 
sity, CA 

Holly L. Gorton, Stanford University, 

Dina F. Mandoli, Stanford University, 

Jeffrey D. Palmer, Stanford University, 

Terri L. Reichert, Stanford University, 

Jeffrey R. Seemann, Stanford University, 

James Shinkle, Stanford University, CA 

William E. Williams, Stanford Univer- 
sity, CA 

Undergraduate Student 
John Schaer, Stanford University, CA 

Clerical and Technical Staff 

Peter A. Angwin, Laboratory Techni- 
cian 7 
John Timothy Ball, Laboratory Techni- 
George W. Bengtson, Jr., Laboratory 

Dorothy Ruth Fischer, Administrative 

Assistant- Accountant 
Glenn A. Ford, Laboratory Manager 
John A. Gamon, Laboratory Technician 
Steven M. Graff, Laboratory Technician 
Richard W. Hart, Mechanical Engineer 
Einar Ingebretsen, Electronic Technician 
Fred Lakin, Technical Illustrator 8 
David W. Lendrum, Technician 9 
Gunilla Malmberg, Laboratory Techni- 
cian 10 
Frank Nicholson, Senior Technician 
Debra Peters, Laboratory Technician 11 
Norma J. Powell, Typist 
Pedro Pulido, Technician 
James M. Tepperman, Laboratory Tech- 
Rudolph Warren, Technician 
Aida Wells, Department Secretary 

7 To September 30, 1980 
8 To May 4, 1981 
9 To September 30, 1980 
10 To January 31, 1981 
u ToMay 31, 1981 

Department of Embryology 

Baltimore, Maryland 

Donald D. Brown 


2 -e ^ m s 

CO 01 . i-K 


s ts H . 5 a 

CO h<n-" CD 

-c a « oj 

co y <p « co 

b p s> "- 1 a 

5 -si c 

CO bD 
•ih <*i 3 C 

c h- 1 a 

CD l-H -H 

8 H 2 <• 

o CD 2 

. >, co v 1 b 
5=5 » I 

T3 >-* CO tn & 


J. H &8 

. CD CD ft 3 
3 )h > bo J} o" 

Xj _CD > o -C -a 

w g CO SC co co 

-* . o i. a 
.a -c co r 2 -3 3 

(£« 3 O 3 J 
.. S ffl ® * 

■K to co en cs 

.Offl CO" S § 

* ft^S I 8 
.S « | g 

S -^ co ccl O to 


, a a « „ 

"" bD 

.3 CO 

<S c "3 o c 

§ a 

■a 3 a 

3 « a "3 

?3 W 

» c B p-h b 

.a; osa ™ 


"l TO li-l ^ 

oW a) - - 

k°H^3 to 

S O CO CO 1=1 

,, "-S C CO ^ _ 

O c! co 3 

a o 

ft "3 .2 "-D 
CO ►* co &c o 

Q £ Q .g pq 


Introduction 109 

Nerve Cell-Synaptic Target Interactions and Nerve Repair 112 

Glial Destruction and S-Cell Synapse Regeneration 113 

Sensory Neuron Regeneration in the Absence of Glia 116 

Accelerated Axonal Degeneration at Elevated Temperatures 120 

Development of Synapses 120 

Localization of Electrical Synapses 121 

Studies on the Cell Surface of Skeletal Muscle Fibers 122 

Monoclonal Antibodies to Acetylcholinesterase 122 

Monoclonal Antibodies to Other Junction Antigens at the Neuromuscular 

Junction 124 

Studies on the Biosynthesis, Secretion, and Extracellular Organization of Fi- 

bronectin by Skeletal Muscle 127 

Extracellular Matrix Organization 129 

Skeletal Muscle Insulin Receptors 132 

Studies on Acetylcholinesterase from Chick and Quail 132 

Purification of Acetylcholinesterase 134 

Studies on Molecular Forms of Acetylcholinesterase in Quail Muscle Cell Cul- 
tures 136 

Dynamics and Metabolism of Cellular Lipids 139 

Kinetics of Phospholipid Transfer Between Vesicles 140 

Studies of Membrane Fusion in Vesicle- Vesicle and Vesicle-Cell Systems ... 141 

Internalization of Plasma Membrane Phospholipids 143 

Cellular Distribution and Metabolism of Fluorescent Phospholipids 144 

Genetic Specification of Cellular Morphology during 

Caenorhabditis elegans Spermatogenesis 147 

An Increase in Intracellular pH Can Induce Spermiogenesis 148 

Few Macromolecular Changes Accompany Spermiogenesis 149 

The Localization of the Major Sperm Protein during Spermatogenesis .... 149 

Genetic Estimation of the Number of Sperm-Specific fer Genes 151 

Somatic Mutation in Antibodies Is Correlated with the Heavy Chain Class 

Switch 152 

Controlling Elements in Maize 158 

The Bronze Locus 159 

The C2 Locus 160 

The Waxy Locus 160 

The Shrunken Locus 164 

Physical Mapping of Chloroplast Gene Mutations: Progress toward DNA-Medi- 

ated Transformation 174 

Transposable Genetic Elements in Drosophila 176 

Molecular Cloning of the Sequences of the White Locus of D. melanogaster . . Ill 

Molecular Analysis of the w DZL Allele 179 

The White-ivory Mutation and Its Revertants 180 

The Molecular Basis of w c Mutability 182 

Molecular Analysis of Hybrid Dysgenesis 184 

The Amplification and Expression of Drosophila Chorion Genes 185 

Structure of Two Chorion Gene Clusters 186 

The Temporal Program of Chorion Gene Expression 188 

The Mechanism of Chorion Gene Amplification 191 

Genetic Analysis of Amplification and Expression 193 

Identification of a Transcriptional Control Region Upstream from the HSV 

Thymidine Kinase Gene 196 

How Does tk mRNA Avoid the Splicing Requirement? 204 

Histone Gene Transcription during Oogenesis 205 

The Dual 5S RNA Gene System in Xenopus 205 

Binding of a Transcription Factor to the Control Region of the 

5S RNA Gene 206 

Construction and Quantitative Transcription of SomaticrOocyte Hybrid 5S 

RNA Genes 208 

A Stable 5S DN A Transcription Complex Can Be Assembled in vitro .... 210 
Chromatin Transcription Retains Developmental Control of Xenopus 5S RNA 

Genes in vitro 212 

5S Transcription Factors in Oocytes and Somatic Cells Have Similar 

Properties 214 

Regulation of Synthesis of the 5 S RNA Transcription Factor 216 

Establishment of the Developmental Control of 5S RNA Genes in Xenopus 

Embryos 216 

The Collection of Human Embryos 217 

Developmental Stages in Human Embryos 217 

Development of the Nervous System 218 

The Developing Skeletal System 218 

Other Organs 220 

Visitors 220 

Staff Activities 220 

Bibliography 222 

Personnel 225 


It seems proper that the mission of 
an embryology department should fo- 
cus toward the sources of some of so- 
ciety's greatest health problems. Per- 
haps five percent of newborn infants 
have congenital defects, most of them 
genetic in basis. Sterility and the con- 
trol of fertility are high on any list of 
human concerns, as are the early men- 
tal and physical development of chil- 
dren. Cancer is often described as 
growth and development out of control. 
All diseases have a developmental com- 
ponent to the extent that they attack 
preferentially certain tissues in the 
body, occur predominantly in one sex, 
and appear at a predictable age. The 
life expectancy of a species and the 
length of gestation are both dependent 
on genes and development. To be sure, 
we cannot precisely measure in finan- 
cial or human terms the burden that 
congenital defects, developmental dis- 
eases, and aging add to society. But the 
alleviation of just one of the many dis- 
orders in these categories would pay 
back our entire national investment in 
biological research. The benefits to so- 
ciety would be perpetual. 

Given these pressing human needs, 
why do we study the nervous system of 
the leech, the eggs of a frog, the sperm 
of a worm, the genes of fruit flies and 
corn, the muscle of chicken and quail? 
We have selected these animals and 
plants because each provides some 
special advantage for studying a par- 
ticular biological problem. The nervous 
system of a leech, unlike that of the hu- 
man, is simple and accessible, and 
makes it possible for us to study the re- 
generation and connections of individ- 
ual nerve cells. The eggs of a frog are so 
large that genes, proteins, and other 
molecules can be introduced within 

them easily. When certain genes mu- 
tate in corn, the event can be detected 
and in some cases understood just by 
looking at a mutated kernel. C. elegans 
and Drosophila have among the 
shortest life cycles of any animals 
studied today. This permits identifica- 
tion and characterization of genetic 
mutations that affect some complex 
function such as fertilization or egg- 
shell formation. Inherent in these stud- 
ies is the assumption that what we 
learn will apply to analogous processes 
in humans; there is considerable prece- 
dent to support this assumption. 

Nerves in humans and leeches are 
ensheathed by glial cells, including 
Schwann cells, whose function remains 
enigmatic. One theory holds that this 
cellular envelope not only guides an 
axon to the tissue it will innervate but 
also may be required to achieve neural 
connections. Muller and his colleagues 
are able to observe the regeneration of 
an individual severed leech neuron as it 
retraces its path to its original target 
tissue, where it then makes an electri- 
cal connection. By selectively destroy- 
ing the glial cell sheath, Muller et al. 
have determined that accurate regen- 
eration of the neuron can proceed in the 
absence of glia. 

The preparation of monoclonal anti- 
bodies directed against surface mole- 
cules of muscle fibers provides Fam- 
brough and his colleagues with a 
wealth of probes enabling them to de- 
tect a variety of determinants in mus- 
cle membranes. Antibodies against the 
enzyme acetylcholinesterase show that 
a patient with "myasthenia syndrome 
with esterase deficiency" has no detec- 
table enzyme at neuromuscular junc- 
tions. This is the second neuromuscular 
disorder whose molecular basis has 




been elucidated by Fambrough and his 
associates. A few years ago they 
showed that patients with myasthenia 
gravis have decreased amounts of ace- 
tylcholine receptors at neuromuscular 
junctions. A major goal of this group 
has been to explain the interaction that 
occurs when nerve innervates muscle. 
Receptors increase at the neuromus- 
cular junction. An important consti- 
tuent is the basal lamina. By using 
monoclonal antibodies to components 
of the lamina, Fambrough et al. have 
observed that the basal lamina in- 
creases when muscle is innervated. 
Newly formed receptor aggregates may 
thus be associated with newly synthe- 
sized basal lamina. 

The importance of the enzyme acetyl- 
cholinesterase makes its synthesis, cell 
location, and general characteristics of 
special interest. Rotundo and Emmer- 
ling are studying the enzyme's molec- 
ular structure under different physio- 
logical conditions in nerve and muscle 
of the chicken and quail. 

Pagano and his associates are devis- 
ing methods to study the composition, 
metabolism, and organization of lipids 
in cell membranes. Their goals resem- 
ble those of Fambrough, whose empha- 
sis is on proteins. The methods needed 
for lipid analysis, however, are very dif- 
ferent from those used for proteins. Li- 
pids are labeled isotopically or with 
fluorescent groups and are introduced 
into membranes by fusion. This 
method of tagging a molecule avoids la- 
beling of the cell pool and introduces 
the molecule preformed into its meta- 
bolic pathway. Pagano is especially in- 
terested in determining the fate of such 
molecules in cells. He and his col- 
leagues are now defining the variables 
that influence the behavior of lipids in 

The motile sperm of the nematode C. 
elegans can be induced to extend their 
pseudopods in vitro. Ward has shown 
that this morphological change occurs 
by rearrangement of preexisting mac- 
romolecules rather than by synthesis of 

new components. Extensive surface re- 
arrangements accompany this change. 
Ward and his colleagues estimate that 
there are about 30 total genes in C. 
elegans that specifically affect sperm 
fertility. They have already identified 
mutants in 16 of them. They are devis- 
ing biochemical and cytological assays 
to probe these sperm-specific mutants. 

During the past 15 years a contro- 
versy has raged concerning the mecha- 
nism whereby an organism can produce 
a seemingly unlimited variety of anti- 
body molecules. Recombinant DNA 
and monoclonal antibody methods 
have revealed that there is not just one 
mechanism, but several. Gearhart pre- 
sents convincing evidence for one of 
these— the somatic diversification of 
antibody genes. The first response to 
an antigen causes an immune cell to ex- 
press one of a small number of germline 
heavy-chain genes (encoding a protein 
termed IgM). The cell then switches to 
production of proteins encoded by 
other heavy-chain genes (termed IgG 
or IgA). This "heavy -chain switch" not 
only involves a second genetic transpo- 
sition but also appears to result in 
mutations within that region of the an- 
tibody gene encoding the antigen-bind- 
ing part of the molecule. 

It is now some 25 years since Bar- 
bara McClintock discovered that ge- 
netic elements in maize can transpose 
in the genome, affecting genes they 
move near or within. With the advent 
of modern molecular genetics it has 
become possible to study these genetic 
events in great detail and to isolate the 
movable elements. The importance of 
transposable elements in eukaryotic bi- 
ology becomes more evident daily. 
Transpositions appear responsible for 
the majority of spontaneous muta- 
tions. Viruses that integrate into euka- 
ryotic genomes are analogous in struc- 
ture and behavior to these elements. 
Directed transpositions such as those 
that occur in the antibody genes are 
now well documented. 

There are two reasons why movable 



genetic elements intrigue a develop- 
mental biologist. First, they cause 
mutations and thus provide a rich 
source of alleles with altered gene func- 
tion. Second, their special ability to 
move from one genetic location to an- 
other invites a detailed explanation in 
molecular terms, and a solution might 
itself suggest a powerful tool for direct- 
ing genetic change. 

Fedoroff and her colleagues have con- 
centrated on four genes in maize known 
to have mutant alleles carrying trans- 
posable elements. Two of these (Bronze 
and C2) encode enzymes on the path- 
way of pigment biosynthesis, while the 
other two {Shrunken and Waxy) encode 
enzymes of carbohydrate synthesis. 
Fedoroff et al. seek to describe the 
changes in the gene caused by these 
genetic elements. Such an analysis re- 
quires ultimately the isolation and 
characterization of wild-type and mu- 
tant genes. Three mutations that affect 
the Shrunken gene have been partly 
characterized as chromosome rear- 
rangements. One of these may be an in- 
sertion within an intervening sequence 
of the gene. 

Transposable elements in Drosophila 
have a different history from those in 
maize. Moderately repetitive DNA se- 
quences dispersed about the genome of 
all plants and animals were discovered 
originally by R. Britten and his col- 
leagues at the Carnegie Department of 
Terrestrial Magnetism in 1968. Later, 
DNA cloning made possible the discov- 
ery that even closely related flies had 
these moderately repetitive DNAs at 
different locations in their genomes. 
Rubin and his colleagues at the Depart- 
ment of Embryology, who had been in- 
volved in these discoveries, then 
sought a gene or gene complex that was 
altered by a transposable element. In 
their report, they describe such a 
system, one which uses one of the best 
known genetic complexes— the white 
locus. Certain mutants affecting this 
locus were predicted and then shown 
by Rubin and his associates to be due 

to the insertion of transposable 
elements. An especially powerful tool 
for studying transposable elements is a 
phenomenon in Drosophila called 
"hybrid dysgenesis." This appears to 
be caused by crossing one strain con- 
taining many copies of a transposable 
element in its genome with another 
that does not. 

As part of their specialized expres- 
sion, a group of genes encoding egg- 
shell proteins in Drosophila are 
amplified just before they are ex- 
pressed. Powerful methods of molecu- 
lar genetics have enabled Spradling 
and his colleagues to map the regions 
along the genome involved in this 
amplification, locate the genes, and 
characterize the RNA transcripts from 
these genes. Using cloned fragments of 
this region, they have found previously 
undetected genes that appear to be 
part of this genetic control system. 
Mutants that rearrange regions within 
the amplified portion alter the amplifi- 
cation process. Spradling and associ- 
ates hope eventually to map the origin 
of this DNA amplification. 

McKnight has carried out the first 
truly systematic analysis of DNA con- 
trol signals in and around a protein-en- 
coding gene. By injecting mutants of 
the herpes simplex thymidine kinase 
gene into Xenopus oocyte nuclei, he has 
delimited a region required for efficient 
transcription of the gene, between 40 
and 100 nucleotides upstream from the 
start site of the gene. This region is 
capable of directing transcription ini- 
tiation at a fixed distance even when 
the gene and its immediate flanking se- 
quence have been deleted and replaced 
with plasmid DNA. It is clear from 
these and other studies that accurate 
initiation of genes encoding proteins, 
i.e., genes transcribed by RNA polym- 
erase form II, require signals in the 
flanking region upstream from the 
start site. This contrasts with the in- 
ternal' control region characteristic of 
genes transcribed by RNA polymerase 
form III, such as the 5S RNA genes. 



Attention in my own lab focuses on 
the developmental control of the dual 
5S RNA gene system in Xenopus. Bo- 
genhagen and Wormington have 
shown that transcription complexes 
formed between a 5S RNA gene and 
cellular extracts in y,itro behave very 
much like chromatin isolated from 
nuclei. Transcription complexes formed 
both in vitro and in chromatin are ex- 
ceedingly stable. It is even possible to 
show that an inactive transcription 
complex is stable. This stability per- 
mits a gene to be transcribed many 
times without dissociating from the 
factors that regulate it, and it provides 
a model for the maintenance of the dif- 
ferential state. 

Our intellectual climate is enriched 
each year by visitors on sabbatical 
leaves from other laboratories. We 
were fortunate to be joined during all or 
part of this year by Joseph Gall of Yale 
University, Laurens Mets of Case Wes- 
tern Reserve, and Kenneth Longmuir 
of the University of California at Ir- 
vine. In addition to such visitors as 
well as postdoctoral students, we ap- 
point one Staff Associate each year. 
This position provides talented young 

investigators with three years of unin- 
terrupted independent research. Gear- 
hart, McKnight, and Rotundo hold 
these positions at present. 

James Blackwell retired this June af- 
ter 19 years of dedicated service. His 
work was marked by quiet efficiency 
and dependability. 

The research of Brown, Gearhart, 
Fambrough, McKnight, Muller, Pa- 
gano, Rubin, Spradling, and Ward has 
benefited from grant support by the 
National Institutes of Health. Fedoroff 
and Rotundo were supported by grants 
from the National Science Foundation, 
and Fedoroff receives support from the 
U.S. Department of Agriculture. Fam- 
brough receives additional support 
from the Muscular Dystrophy Associa- 
tion. Pagano has a grant from the 
Whitehall Foundation, and Rubin is 
supported by an American Cancer 
Society grant. Departmental research 
has been assisted greatly by grants 
from the Commonwealth Fund, the 
Culpeper Foundation, and the Fleisch- 
mann Foundation, as well as by a 
Biomedical Research Support Grant 
from the National Institutes of Health. 
Donald D. Brown 



K. J. Muller, E. J. Elliott, A. Mason, and B. E. Thomas 

The functioning of the nervous sys- 
tem relies on the formation and main- 
tenance of a complex and highly spe- 
cific set of connections between nerve 
cells, or neurons. These connections, 
called synapses, are not necessarily fixed 

or static throughout the life of the 
animal; for some neurons, synapses 
may be regenerated to restore function 
after nerve damage. We have been 
studying the functioning and repair of 
neurons and their synapses in the me- 



dicinal leech in order to understand 
nerve signaling and regeneration at the 
level of single cells. 

The central nervous system of the 
leech is particularly favorable for study 
because it contains easily identifiable 
neurons in relatively few numbers. Fur- 
thermore, the properties of individual 
neurons and their synapses are strik- 
ingly like those in diverse systems, in- 
cluding the human brain. Some 
synapses are chemical, where doses of 
transmitter molecules released by one 
cell diffuse a short distance to the next 
cell to activate receptor molecules in its 
membrane; other synapses are elec- 
trical and carry currents directly from 
one neuron to the next. In addition, as 
in higher animals, the neurons, nerve 
fibers (axons), and synapses in the 
leech are ensheathed by nonneuronal 
glial cells. It has been shown with 
physiological and anatomical tech- 
niques that leech neurons can accu- 
rately regenerate severed axons and 
form synapses precisely with their nor- 
mal targets. Depending upon the par- 
ticular neuron under study and the site 
of injury to its axon, it takes about a 
month for the nervous system to re- 
cover. During this period, the portion 
of axon cut from the cell body survives 
intact, as do severed pieces of glial cell, 
including those with no nucleus. 

In other systems, glial cells, in- 
cluding Schwann cells, have been 
thought to guide growing axons. This 
year we have selectively destroyed 
leech glial cells that ensheathe axons 
running between ganglia in order to 
test whether these large cells are re- 
quired for accurate synapse regenera- 
tion. Last year we reported on prelimi- 
nary experiments performed in vitro; 
the recent investigations were per- 
formed in the animal. We found that 
synapses form with normal or nearly 
normal reliability and accuracy. 

We have also examined the possibil- 
ity that glia maintain severed distal 
stumps of axons and act to isolate ax- 
ons from one another, thus preventing 

inappropriate growth or sprouting. Our 
results suggest that the large glia may 
play a limited role. 

Our other research this year included 
continued investigation of how severed 
axonal stumps survive and the stump's 
possible role in regeneration, detailed 
examination of the pattern of multiple 
contacts between cells, particularly 
between sensory and motor neurons, 
and the beginning of a study of the ap- 
pearance of synapses in the developing 
nervous system. 

Glial Destruction and S-Cell 
Synapse Regeneration 

E. J. Elliott and K. J. Muller 

One particular neuron, the S-cell, has 
been especially useful for studies of 
synapse regeneration. Because its axon 
is the largest in the leech connectives 
(the axon bundles that link adjacent 
ganglia), the axon is both anatomically 
and electrophysiologically distinctive. 
Moreover, the single S-cell in each seg- 
mental ganglion forms, at the tip of 
each of its two axonal branches, an elec- 
trical synapse with its homologues. 
Synapses between S-cells occur approx- 
imately midway between ganglia in the 
connectives, a region that is much 
easier to analyze than is the complex 
neuropile within ganglia where most 
synapses are made. We have previously 
found that if we injure an S-cell by cut- 
ting or crushing the connectives, it will 
grow a thin, new axon along its severed 
stump, which survives physiologically 
and morphologically intact. The grow- 
ing axon forms an electrical synapse 
specifically with its S-cell target at the 
connective midpoint, whereupon it 
stops growing. Sometimes as the 
S-axon grows along the distal stump it 
forms an electrical synapse with the 
stump. Because the stump remains 
connected to the target, the stump 
links the regenerating neuron with its 
target. Function is thereby restored 
unusually rapidly. 



In a series of earlier experiments, we 
found that when the target S-cell is se- 
lectively destroyed by an intracellular 
injection of protease, the regenerating 
neuron nonetheless grows in a normal 
fashion toward the original site of 
synapse (near the connective mid- 
point), where the axon stops without 
forming alternative electrical synapses 
with other neurons (Year Book 78, 79). 

Glial cells, including Schwann cells, 
have long been thought to be important 
in guiding regenerating neurons. In 
light of the surprisingly limited role of 
the target S-cell in synapse regenera- 
tion, it was of interest to apply a 
similar approach in investigating the 
role of the large glial cells of the connec- 
tive, which ensheathe the growing axon 
and the site of synapse. Two lateral 
connectives and a smaller, medial con- 

nective link adjacent ganglia; a single 
large glial cell ensheathes the axons of 
each lateral connective and one or both 
glia ensheathe the medial connective, 
called Faivre's nerve. To be certain of 
destroying the glial sheath in Faivre's 
nerve, we therefore injected both lat- 
eral glial cells with sufficient protease 
to destroy them selectively. In six 
preparations examined one week after 
protease injection, the glial cells were 
destroyed (Fig. 1). At this point, the 
S-cell axons were severed by crushing 
the connectives with fine forceps or by 
cutting through the lateral connective 
and Faivre's nerve with iridectomy 
scissors. We examined 33 preparations 
at times ranging from 9 days to 24 
weeks after crushing the connectives, 
and four preparations were examined 
up to 25 weeks after cutting. The re- 




Fig. 1. Glial cells that ensheathe the medial connective (Faivre's nerve) were selectively 
destroyed by intracellular injection of protease. Both glial cells of the lateral connectives, which also 
ensheathe Faivre's nerve, were injected with protease, and by seven days only scattered glial debris 
remained. In cross sections of Faivre's nerve (A), the locations of nuclei of two microglial cells (M) 
and one S-cell axon (S) are indicated. Enlargement (B) shows microglial cell near extracellular debris 
consisting of multivesicular bodies and (in C) bundles of intermediate filaments (IF). 



suits were nearly indistinguishable 
from those obtained in the presence of 
the glial cell. The S-cell regenerated its 
synapse with an 80% success rate after 
its axon had been severed by crushing 
the connective. Each axon grew along 
its original pathway, ordinarily marked 
by the severed stump, to synapse spe- 
cifically with its S-cell target near the 
connective midpoint (Fig. 2). In five 
cases, the regenerating neuron syn- 
apsed with its severed stump before 
regeneration was complete, forming a 
basket of processes around the stump, 
which acted as a weak electrical splice 
or bridge via its maintained connection 
with the target (Fig. 3). Impulses did 
not conduct between S-cells that were 
coupled through the stump. Evidence 
indicates that neurons continued to 

grow toward the target after synapsing 
with the distal stump. 

Previous experiments have shown 
that when the glial cell is present the 
distal stump survives until regenera- 
tion is complete, degenerating during 
the second month following the lesion. 
It seems that regeneration itself trig- 
gers degeneration for if regeneration 
fails, either because the initial lesion is 
too severe or for other, unknown rea- 
sons, then the severed distal stump can 
survive for months. Experiments in 
which target S-cells were killed by in- 
tracellular injection of protease have 
shown that the distal stump does not 
rely for survival upon the passage of 
metabolites across the electrical syn- 
apse made with the target S-cell. It has 
been postulated that in invertebrates 



Fig. 2. Regeneration of the electrical synapse between S-cells after destruction of the connective 
glial cells. The glial cells of the lateral connectives were injected with protease within the animal, and 
one week later the axon of the S-cell in ganglion 11 was severed by crushing the connectives, as indi- 
cated by the horizontal bar in the schematic diagram. Four weeks later, horseradish peroxidase in- 
jections indicated that the severed S-cell axon had grown anteriorly along its distal stump and recon- 
nected midway between ganglia with the S-cell in ganglion 10. Depolarizing currents delivered 
through the recording microelectrode into either cell elicited impulses that propagated with little 
delay across the regenerated synapse (upper traces; current on third trace). Hyperpolarizing cur- 
rents spread across the synapse with a normal, approximately tenfold, attentuation (lower traces; 
current on third trace). 




>0 um 

Fig. 3. The regenerating S-cell axon can form 
a basket of processes around the severed distal 
stump and synapse with it, thereby linking 
S-cells in adjacent ganglia before regeneration is 
complete. While synapsing with the stump can 
normally occur, here the connective glial cells 
were destroyed with protease and one week later 
the axon of the S-cell in ganglion 10 was severed 
by crushing the connectives. By 27 days, at left 
(A), the axon had grown along its stump toward 
the target S-cell in ganglion 9, which did not 
grow. The basket of processes, shown at higher 
magnification at the right (B), were seen in the 
electron microscope to be associated with the 
severed distal stump. 

the ensheathing glial cell maintains 
severed axonal stumps. Our results in- 
dicate that severed distal stumps can 
survive for normal periods if the glial 
cell has been destroyed, but in some 
cases the stump may prematurely de- 
generate. It may be significant that 
once the large glial cells are destroyed, 
small macrophage-like cells called 
"microglia" proliferate within the 
nerve cord and eventually ensheathe 
portions of the naked axons. The small 


cells initially contain lysosomal inclu- 
sions and extensive rough endoplasmic 
reticula, and they eventually acquire 
some intermediate filaments. While 
these new cells would not be expected 
to be able to mark the original axonal 
pathway, they might act as metabolic 
replacements for the large glial cells. 
Experiments are in progress to trace 
the origin of the microglia and to deter- 
mine whether such cells that appear 
after injury become linked to the large 
glial cells by specialized junctions. 

The regenerated electrical synapse 
between S-cells can be detected elec- 
trophysiologically by showing that 
current can again pass from one S-cell 
to the next (Fig. 2) and that the two 
cells are in direct contact. A well-estab- 
lished synapse allows nerve impulses 
to travel in either direction between 
cells, but in the early stages of regener- 
ation impulses may propagate in only 
one direction across the synapse or 
may be intermittent. We have found 
similar stages in the formation of 
synaptic transmission, whether the 
large glial cells are present or not, in- 
cluding one-way transmission of im- 
pulses from the distal stump into the 
fine, regenerating axon. It is thought 
that one-way transmission occurs be- 
cause excitable channels in the regener- 
ating axon are unable to supply enough 
current to the larger stump to excite it. 
As the regenerating axon grows in 
caliber and forms a more extensive 
synapse, impulses can travel in both di- 
rections across the synapse and full 
function is restored. 

Sensory Neuron Regeneration in 
the Absence of Glia 

E. J. Elliott and K. J. Muller 

We have studied the effect of glial 
killing on regeneration for a second 
group of neurons— the mechanosensory 
neurons. Like S-cells, sensory neurons 
can regenerate severed axons and can 
accurately reinnervate their correct 



target cells. The frequency of accurate 
regeneration of sensory cell axons is 
much lower, however, than in S-cells. It 
is not known whether this indicates a 
qualitative difference in the mechanism 
of regeneration. A second difference 
between S-cells and the mechanosen- 
sory neurons is that S-cell axons form 
electrical synapses with their targets, 
while the sensory neurons form chem- 
ical as well as electrical synapses in ad- 
jacent ganglia. Whether the different 
types of synapses regenerate by differ- 
ent mechanisms is unknown. 

The sensory neurons selected for 
study were those responding to moder- 
ate pressure (P-cells) and those re- 
sponding to noxious stimuli (N-cells). 
These cells occur in pairs, one on each 
side of a ganglion, and send their axons 
down the ipsilateral connectives to ad- 
jacent ganglia. In the present experi- 
ments, the glial cell of one of the two 
lateral connectives was killed by in- 
tracellular injection of protease. The 
other lateral connective served as a 
control. A week later, after most of 
the debris from the killed glial cell had 
disappeared, both connectives were 
crushed or cut in a second operation, 
severing all of their axons. At various 
times from 3 to 20 weeks after injury, 
the connectives and the adjacent gan- 
glia were dissected from the animal for 
final examination. 

Regeneration of P and N sensory ax- 
ons in both connectives was monitored 
electrophysiologically by stimulating 
the sensory cell bodies and recording 
responses in the L motor neuron in the 
adjacent ganglion, one of their normal 
synaptic targets. Successful regenera- 
tion resulted in the reappearance of a 
monosynaptic excitatory postsynaptic 
potential in the L motoneuron follow- 
ing an action potential in the sensory 
P-cell or N-cell (Fig. 4). Successful 
regeneration was found to occur both 
for glial-ensheathed axons and nonen- 
sheathed axons. While the frequency of 
regeneration for glial-ensheathed ax- 
ons was greater (25%, or 15 of 60 cells 

tested) than that for nonensheathed 
axons (16%, or 10 of 64 cells tested), 
this difference is probably not signifi- 

After each electrophysiological test, 
one pair of sensory neurons was in- 
jected with horseradish peroxidase 
(HRP) as a marker, and the preparation 
was placed in modified L-15 culture 
medium for three days to allow com- 
plete diffusion of the HRP within the 
processes of the injected cells. The 
preparations were then fixed, stained, 
and viewed with the light microscope. 

The morphological appearance of the 
injected cells was generally consistent 
with the physiological experiments. In 
all cases where functional regeneration 
had been shown electrophysiologically, 
the regenerating axon had grown 
across the lesion and into the next 
ganglion (Fig. 5). In addition, in many 
cases where no functional regeneration 
was detected, either the axon did not 
grow across the site of injury or its 
growth stopped short of the next gan- 
glion. However, in some cases where 
functional regeneration had not oc- 
curred, the axon had grown to the next 
ganglion. It is possible in these cases 
either that the axon was growing in a 
correct direction and had not yet made 
contact with the L motor neuron at the 
time of the experiment, or that the ax- 
on had made connections with other 
neurons, including unusual targets, in 
the ganglion. A search for such unusual 
or aberrant connections was made in a 
number of preparations by stimulating 
the sensory neuron and recording re- 
sponses in the adjacent ganglion from 
neurons near the cell body of the L 
motor neuron and from other motor 
neurons. No synaptic potentials in 
these other neurons were observed. 

The morphological experiments also 
revealed a marked difference in the 
growth pattern of axons whose glial 
sheath had been destroyed. While ax- 
ons on both control and protease- 
treated sides sprouted in the region of 
the crush or cut, axons on the protease- 




L - 



Fig. 4. Oscilloscope records of synaptic potentials recorded in the ipsilateral L motor neuron of 
the adjacent posterior ganglion after stimulating sensory P- and N-cells. Synaptic potentials from 
regenerated synapses on the control side (A) are essentially indistinguishable from those on the 
protease-injected side (B). The traces for regenerated P synapses are from the same preparation, 
eight weeks after a crush near the posterior ganglion. The traces for the regenerated N synapses are 
from preparations eight weeks (control) and six weeks (protease-injected) after a crush near the 
posterior ganglion. All synaptic potentials were recorded in leech Ringer with a Ca concentration of 
7.5 mM. 

treated side sprouted, often profusely, 
in regions proximal to the lesion (see 
Fig. 5). In some cases, the sprouts prox- 
imal to the lesion resulted from pro- 
cesses originating at the lesion and 
growing toward the ganglion of the sen- 
sory cell body, away from the ganglion 
of the target neuron. 

These results present two principal 
findings: (1) axons in the leech can re- 
generate accurately in the absence of 
the ensheathing connective glial cell, 
and (2) eliminating the glial cell does af- 
fect the pattern of axon growth. Fur- 

ther interpretation of these results 
may be complicated by at least three 

First, in these experiments the final 
region of growth of the regenerating 
axons, including the region of synapse 
with the target neuron in the next gan- 
glion, is still glial-ensheathed by the 
neuropil glia. It is possible that growth 
of the axon along the connective is 
largely random and that the axon is 
guided in its final stages of growth by 
the glial cells of the ganglion contain- 
ing the target neuron. It would be 



Fig. 5. Camera lucida drawing of a pair of P 
sensory neurons that have both regenerated ax- 
ons across a crush (at arrow) made four months 
earlier. The axon for which the ensheathing glial 
cell was destroyed by protease injection (P) has 
sprouted profusely proximal to the lesion, while 
the axon on the control side (C) shows no 
sprouting in this region. Sprouting probably oc- 
curred earlier on both sides at the lesion, and 
some sprouts then retracted after a functional 
synapse was established. Physiological experi- 
ments indicated that both cells had made cor- 
rect synaptic connections with the ipsilateral L 
motor neuron, a proper target. Arborization of 
P-cell processes in the ganglia is more extensive 
than shown here (anterior at top). 

technically difficult to destroy those 
glial cells by protease injection. How- 
ever, experiments described above 
show that killing the glial cell en- 
sheathing the synaptic region of the 

S-cell does not prevent accurate regen- 
eration of an S-cell synapse. 

Second, the pattern of sprouting 
seen for axons not ensheathed by glia 
may be due to the presence of protease, 
rather than to the destruction of a glial 
cell per se. To test this, protease was 
injected extracellularly under the con- 
nective tissue sheath of the connective; 
from three to five weeks later, sensory 
neurons were injected with HRP to 
check for sprouting of the axons on the 
protease-injected side. No sprouting 
was seen. 

In addition, it is likely that destruc- 
tion of the glial cell itself may prompt 
sensory neuron sprouting. Connective 
glial cells were injected with protease 
and the connectives were not crushed; 
from three to thirty weeks later, sen- 
sory neurons were injected with HRP. 
In only a few cases did axons sprout. 

Third, while the large, ensheathing 
glial cells have been destroyed, there is 
a proliferation of the small, macro- 
phage-like microglial cells. (See dis- 
cussion in previous section.) Future ex- 
periments aim to quantify the increase 
in microglia after injury and to deter- 
mine whether the microglia themselves 
are important for accurate axon regen- 

In summary, these experiments sug- 
gest that the glial cell does provide 
orienting information to growing ax- 
ons in the leech connective, since the 
growth pattern is altered in the ab- 
sence of the glial cell. However, the 
glial cell is not the only source of tactic 
guidance, since axon growth in the cor- 
rect direction does still occur in the 
absence of the glial cell. Another possi- 
ble source of tactic guidance is the 
distal stump of the severed axon. In 
the case of the S-cell, the distal axon 
stump is known to survive for months, 
and the regenerating proximal axon 
process grows along the stump. Pre- 
liminary experiments indicate that the 
distal stumps in the connectives of sen- 
sory neurons survive for at least four 
weeks. Experiments are in progress to 



determine whether regenerating sen- 
sory axons grow along their distal 

Accelerated Axonal Degeneration 
at Elevated Temperatures 

A. Mason and K. J. Mutter 

One remarkable difference between 
severed axons in the leech and those in 
mammals is that mammalian axons de- 
generate within days, while distal 
stumps in the leech and many other in- 
vertebrates can survive for months de- 
tached from their cell bodies. This dif- 
ference might be important in neuronal 
regenerative capabilities. Because sev- 
ered axons in hibernating ground 
squirrels can for months survive and 
continue to transmit at synapses, it 
seemed possible that temperature 
might determine the rate of degenera- 
tion and that leech axons would degen- 
erate rapidly at elevated temperatures. 
While the medicinal leech Hirudo 
medicinalis does not fare well above 
room temperature, the giant Amazon 
leech Haementeria ghilianii can live 
above 30 °C. 

We cut the ventral nerve cord near a 
mid-body ganglion in specimens of H. 
ghilianii (obtained from the laboratory 
of Dr. G. Stent at Berkeley or raised in 
our own laboratory). After periods 
ranging from one to five weeks we ex- 
amined the distal piece of severed axon 
(distal stump) of the S-cell, which in 
Haementeria resembles that in Hirudo 
in its morphology and physiology. In 
most cases the adjacent S-cell that re- 
mained electrically coupled to the dis- 
tal stump at the normal synapse was 
injected with the marker horseradish 
peroxidase to permit unequivocal iden- 
tification of the severed stump in the 
electron microscope. While more than 
half (15 of 26) of the axons had degener- 
ated or were degenerating within four 
weeks in animals maintained con- 
tinuously at 31 °C, no axons had 
degenerated in twelve animals main- 

tained at room temperature (22 °C) 
when examined at up to three months. 
The survival of some distal stumps of 
S-cell axons at 31 °C suggests that fac- 
tors other than temperature may ac- 
count for some of the long-term sur- 
vival of severed axons in the leech, but 
clearly temperature is an important 

Development of Synapses 

K. J. Mutter and B. E. Thomas 

While regenerating neurons in the 
leech can synapse specifically with 
those targets with which they connect 
during development, we do not know 
how developing neurons differ from re- 
generating neurons in the cellular en- 
vironment wherein they grow, in the 
spatial relationships with their targets, 
and in the steps taken toward making 
synapses. It is technically difficult to 
study development in the medicinal 
leech Hirudo medicinalis, but the 
glossiphoniid leeches Helobdella trise- 
rialis and Haementeria ghilianii have a 
straightforward development, can be 
raised in the laboratory, and have in- 
dividual neurons strikingly like those 
in Hirudo. In collaboration with Dr. 
David Weisblat at Berkeley, we have 
begun to examine with the electron 
microscope the process of synapse for- 
mation in embryonic leeches. It has 
been necessary to modify standard fix- 
ation procedures and fix in a fresh mix- 
ture of glutaraldehyde and osmium to 
retain the delicate structure of the 
developing neurons. This procedure 
should nonetheless be compatible with 
horseradish peroxidase staining of sin- 
gle neurons if tissue is stained before 
fixation (see below). Our preliminary 
results indicate that structures iden- 
tifiable as synapses are forming when 
the embryo begins to hatch from its 
vitelline membrane (after the nerve 
cord has fused and abundant fine 
neurites have been elaborated in the 
neuropil at the center of the ganglion). 



At this early stage in Helobdella, it is 
possible to recognize the profiles of 
S-cell axons in Faivre's nerve, a feature 
that should be useful in determining 
the configurations of axons at their 
earliest synaptic contact. 

Localization of 
Electrical Synapses 

K. J. Muller and B. E. Thomas 

If current can be passed from one cell 
to a separate cell by means of intracel- 
lular microelectrodes, the cells are gen- 
erally considered to make an electrical 
synapse with one another. At many 
electrical synapses, small dye mole- 
cules of molecular weight 1000 daltons 
or less can pass freely between elec- 
trically coupled cells. The exact loca- 
tion of the synapse may be difficult to 
detect, however, and demonstration in 
the electron microscope of a gap junc- 
tion, the morphological correlate of an 
electrical synapse, relies upon fortui- 

tous sectioning and the identification 
of each cell. We have been able to dem- 
onstrate the sites of electrical synapse 
between certain neurons by injecting 
two different markers in combination 
into a single cell. One is the fluorescent 
dye Lucifer Yellow (Mr ~ 450 daltons), 
which crosses many nonrectifying elec- 
trical junctions (junctions at which cur- 
rents can pass in either direction); the 
other is horseradish peroxidase (HRP, 
Mr -40,000 daltons), which is ordi- 
narily too large to cross at electrical 
junctions. The dense reaction product 
catalyzed by HRP can be formed be- 
fore fixation, and the sites of dye cou- 
pling between the injected cell and 
other cells in the nervous system can 
be determined with the aid of a fluor- 
escence microscope. Tissue can sub- 
sequently be prepared for electron 
microscopy. The suitability of the tech- 
nique has been confirmed by localizing 
the electrical synapse between S-cells 
(Fig. 6), and has already proven useful 
in mapping small networks of neurons. 

Fig. 6. Site of an electrical synapse in the leech interganglionic connective, revealed by combined 
injection of Lucifer Yellow dye (LY) and horseradish peroxidase (HRP) into a single cell. The S-inter- 
neuron whose axon is stained darkly on the left was injected with HRP and, two hours later, with 
LY. The site of the barrier to HRP but not to LY is indicated by the arrow midway between ganglia 
and is the site of the electrical synapse. LY fluorescence in the axon injected with HRP was obscured 
or quenched by the diaminobenzidine reaction product. 




M. J. Anderson, E. K. Bayne, M. Chiquet, J. M. Gardner, A. K. Student, and D. M. Fambrough, with 
the technical assistance of D. Somerville 

Our focus of research has been upon 
those mechanisms that control the 
expression and organization of cell sur- 
face properties during muscle develop- 
ment and neuromuscular junction for- 
mation. Acetylcholine receptors and 
acetylcholinesterase of skeletal muscle 
participate in these phenomena and, be- 
cause of the availability of highly spe- 
cific and strongly binding probes, have 
been major subjects of research in past 
years. To gain a more-encompassing 
view of the complexity and organiza- 
tion of the cell surface, we have 
generated a large number of mono- 
clonal antibodies that bind to antigenic 
determinants on the surface of muscle 
fibers. Most or all of these antigenic de- 
terminants are unique chemical groups 
on the various cell surface protein mole- 
cules. Each monoclonal antibody, there- 
fore, defines a species of cell surface 
molecule and is useful in determining 
the distribution of the molecule on the 
cell surface. These monoclonal anti- 
bodies are also useful in experiments 
designed to determine the physico- 
chemical properties of the antigenic 
molecules and to explore the sites of 
biosynthesis and the timing of produc- 
tion during myogenesis and synapto- 

Monoclonal Antibodies 
to Acetylcholinesterase 

In collaboration with Dr. Andrew G. 
Engel, Department of Neurology, Mayo 
Clinic, and Dr. Terrone L. Rosenberry, 
Department of Pharmacology, Case 
Western Reserve University, we have 
generated hybridomas that make anti- 
bodies targeted to human red blood cell 
acetylcholinesterase (AChE). Six inde- 
pendently derived hybridomas were 

found. Culture medium conditioned by 
each of the six contained antibodies 
which bound not only to red blood cell 
AChE but also to AChE at the neuro- 
muscular junctions in sections of hu- 
man skeletal muscle. In tests with 
muscle tissue from various laboratory 
animals, we found that each of the anti- 
bodies cross-reacted with monkey, four 
cross-reacted with dog, and three with 
guinea pig. An example of a guinea pig 
neuromuscular junction stained with 
fluorescent antibody is shown in Fig. 7. 
Based upon differences in cross-reac- 
tivity of antibodies with AChE of var- 
ious animal species, it was judged that 
the antibodies defined at least three 
and probably five or six different an- 
tigenic determinants. Each of the an- 
tibodies was shown to bind to and 
thereby change the sedimentation rate 
of human red blood cell AChE (Fig. 8). 
By mixing different antibodies to- 
gether with AChE and studying the 
sedimentation behavior of the resulting 
complexes, five of the antibodies were 
found to differ in their interactions with 
the acetylcholinesterase molecules. 
Thus the set of monoclonal antibodies 
identifies many of the surface features of 
the acetylcholinesterase molecule. This 
point is important for several reasons. 
First, the fact that all of these different 
antibodies react with both red blood 
cell and skeletal muscle acetylcholin- 
esterase shows that there is a high level 
of molecular homology between the en- 
zymes from these two sources. Second, 
the antibodies can serve as probes for 
different parts of the acetylcholines- 
terase molecule. In this connection, it 
was of great interest to determine 
whether acetylcholine molecules oc- 
curred at the neuromuscular junctions 
of a patient with a syndrome known as 
"myasthenic syndrome with esterase 



Fig. 7. Distribution of acetylcholine receptors (left panel) and acetylcholinesterase (right panel) 
at a neuromuscular junction in guinea pig forearm flexor muscle. The muscle in organ culture was in- 
jected with a solution containing both tetramethylrhodamine-labeled a-bungarotoxin (to label 
acetylcholine receptors) and a monoclonal antibody directed against the acetylcholinesterase. After 
removal of unbound ligands, the muscle was injected with fluorescein-labeled anti-mouse IgG. Single 
fibers were teased from the muscle after mild fixation. These were mounted and the fluorescence due 
to each ligand was observed separately. 

deficiency." While it was known that 
esteratic activity at the neuromuscular 
junctions was virtually nonexistent, it 
was not known whether this reflected 
some abnormality in the enzyme or a 
complete lack of the enzyme. Tests 
with biopsy material from a patient 
with the syndrome revealed that none 
of the antigenic determinants defined 
by our monoclonal antibodies existed 
at the patient's neuromuscular junc- 
tions. Thus, it was concluded that the 
patient lacks the AChE molecules in 
toto. A third use of these monoclonal 
antibodies has been in probing the rela- 

tion between the acetylcholinesterases 
of muscle and nervous tissue. Two 
areas of highest concentration of AChE 
in the central nervous system are the 
caudate nucleus and the spinal cord. 
Using autopsy material, we found that 
two of the monoclonal antibodies 
showed little or no reactivity with the 
acetylcholinesterase of these tissues, at 
least not enough to be useful in anti- 
body staining of tissue sections. Fur- 
ther studies tested for effects of the 
monoclonal antibodies on the sedimen- 
tation properties of solubilized human 
brain AChE. Several of the monoclonal 



(Bottom) l0 


Fig. 8. Velocity sedimentation of human red 
blood cell AChE in the presence (closed circles) 
and absence (open circles) of bound antibody. 
Samples containing about 0.02 /xg of purified 
human AChE were incubated for 1 hour at room 
temperature with or without an added 500-fold 
excess of monoclonal antibody. The samples 
were then subjected to velocity sedimentation, 
and fractions from the gradients were analyzed 
for the distribution of AChE. The sedimentation 
velocity of the AChE -antibody complexes was 
much greater than that of the free AChE. 

antibodies at high concentration 
caused some shift in sedimentation 
rate, while others had no effect. Thus it 
appears that the nervous system ace- 
tylcholinesterases are distinct molec- 
ular species. Of particular interest in 
this regard is the fact that one of the 
antibodies, which recognizes the neuro- 
muscular junctional AChE of guinea 
pigs, did not bind to neuronal AChE, as 
judged by this test. Thus, the staining 
of junctional regions, as illustrated in 
Fig. 7, apparently is due to acetylcho- 
linesterase of muscular origin. The 
staining is quite intense, suggesting 
that most or all of the junctional AChE 
is muscular in origin. This has been pre- 
sumed on the basis of several observa- 
tions less direct than the immunologi- 
cal evidence presented here. Further 
studies with the monoclonal antibodies 
are needed to assess quantitatively the 
muscular and neuronal origins of 
AChE at the junctions. 

Monoclonal Antibodies to Other 
Junction Antigens at the 
Neuromuscular Junction 

As reported in Year Book 79, we 
have found that some of our mono- 
clonal antibodies recognize elements of 
the extracellular matrix of skeletal 
muscle which appear concentrated at 
neuromuscular junctions. During the 
past year we have cloned four an- 
tibodies that recognize basal lamina 
antigens at Xenopus neuromuscular 
junctions (see Fig. 9). We have also 
used cultures of Xenopus nerve and 
muscle cells to study the appearance of 
the antigens during muscle maturation 
and their elaboration during innerva- 
tion by motor neurons. On cultured em- 
bryonic myotomal muscle cells, a com- 
plex, ordered array of basal lamina an- 
tigen appears. This consists of a dif- 
fuse background over the entire muscle 
cell surface, along with a scattering of 
discrete plaques and fibrils. In such 
cells, developing in the absence of in- 
nervation, dense aggregates of acetyl- 
choline receptors invariably appear as- 
sociated with corresponding specializa- 
tions of adjacent basal lamina antigens 
(Fig. 10). When embryonic muscle cells 
become innervated by cholinergic neu- 
rons in vitro, there is a progressive 
nerve-induced accumulation of acetyl- 
choline receptors in the developing 
post-synaptic membrane. This accumu- 
lation of receptors is closely paralleled 
by a corresponding deposition of new 
basal lamina associated with the newly 
formed acetylcholine receptor clusters. 
We conclude that the junctional spec- 
ialization of basal lamina is induced by 
the developing motor nerve and ap- 
pears concomitantly with the receptor 
aggregates in the post-synaptic mem- 

Monoclonal antibody 33 recognizes 
an antigen concentrated in the synap- 
tic areas of chicken muscle and is pres- 
ent both as a diffuse layer and as 
patches on cultured chick myotubes 
(Fig. 10). In these respects the antigen 













Fig. 9. Immunofluorescent localization of basal lamina antigen at a Xenopus neuromuscular 
junction. Upper panel (A) shows distribution of tetramethyl-rhodamine-labeled antibody; lower 
panel (B) shows fluorescence due to fluorescein-labeled a-bungarotoxin at the same junction. 

defined by antibody 33 resembles the 
basal lamina antigens of Xenopus mus- 
cle, described above. It has proved 
possible to isolate small quantities of 
33-antigen from embryonic chick skel- 
etal muscle by lengthy extraction of 
the extracellular matrix fraction with 
0.5 M NaCl, 50 mM Tris-HCl buffer, 
pH 7.5. The extract also contains col- 
lagens and other extracellular matrix 
materials. From the extract the 33- 
antigen has been partially purified by 
gel filtration on BioGel A5M, followed 
by dialysis at room temperature against 
0.5 M NaCl, 50 mM Tris. The resulting 
soluble fraction is greatly enriched in 
33-antigen, and the antigen in this frac- 
tion has been analyzed by velocity sedi- 
mentation and by treatment with vari- 
ous enzymes and other reagents. The 
antigen sediments as a set of bands 
with S-values from approximately 9 to 
16, consisting of two prominent peaks 
with a minor peak in between. In the 
presence of 2 M urea the pattern simpli- 
fies to a single peak (Fig. 11). When an- 

tigen is treated with collagenase under 
conditions which completely fragment 
added collagen, the 33-antigen is un- 
altered; trypsinization results in pro- 
duction of small antigenic fragments 
(Fig. 11). The antigen is destroyed by 
denaturation and by reduction with 

Because of some similarities of 
33-antigen to a basal lamina compo- 
nent, laminin, of mammalian muscle, 
we investigated the possibility that 
33-antigen might be a chicken laminin. 
For this study we used antiserum made 
in sheep against mouse laminin to- 
gether with fluorescent anti-sheep IgG 
to determine the distribution of chicken 
laminin in sections of chicken muscle, 
in tissue-cultured chick muscle, and in 
sucrose gradients after velocity sedi- 
mentation of 33-antigen-enriched frac- 
tions. The antiserum did yield immuno- 
fluorescent staining of cells, showing 
the presence of the cross-reacting 
chicken laminin both as a basal lamina 
component of adult muscle and as an 



extracellular material in myogenic cell 
cultures. However, the distribution of 
laminin, defined in this manner, was 
different from the distribution of 33- 

antigen in that laminin was not found 
to be concentrated at neuromuscular 
junctions and was found to form a ma- 
trix morphologically distinguishable 

■v. ' 

Fig. 10. Localization of basal lamina antigens at acetylcholine receptor clusters on cultured em- 
bryonic muscle cells. Phase-contrast and fluorescence micrographs of chick and Xenopus muscle 
cells in culture. A, B: Chick (A) and Xenopus (B) muscle cells stained with tetramethylrhodamine- 
labeled a-bungarotoxin, which reveal discrete clusters of densely packed acetylcholine receptors in 
the sarcolemma. C, D: Same fields as A, B, stained with fluorescein-labeled antibody, which reveal 
local accumulations of basal lamina antigens over each of the receptor clusters. E, F: Phase-contrast 
views of the fields shown above. The bars in E and F represent 30 ^m. 





20 30 


( Bottom ) 


Fig. 11. Velocity sedimentation analysis of 33-antigen and protease-treated antigen. Antigen 
purified as described in the text was incubated with collagenase (closed circles), trypsin (open 
triangles), and without protease (open boxes) for sufficient time to completely fragment added col- 
lagens or fragment all 33-antigen molecules. The samples were then loaded on sucrose gradients con- 
taining 2 M urea, 0.5 M NaCl, 50 mM Tris, pH 7.5, and centrifuged for 16 hours at 32,000 rpm, 4°C, 
in SW 50.1 rotor. Fractions were collected in microtiter wells and assayed for 33-antigen, with 
monoclonal antibody 33 and 125 I-labeled rabbit anti-mouse Fab second antibody. 

from 33-matrix in vitro. The sedimenta- 
tion profile for chicken laminin was dif- 
ferent from that of 33-antigen, although 
similarities in sedimentation suggest 
that the two may have similar molecu- 
lar weights. For these studies we are in- 
debted to Dr. George Martin and Dr. 
Hynda Kleinman, who generously pro- 
vided the anti-laminin serum. 

Studies on the Biosynthesis, Secretion, 

and Extracellular Organization of 

Fibronectin by Skeletal Muscle 

The view that the extracellular ma- 
trix is the product of connective tissue 
cells (e.g., fibroblasts, chondrocytes, 
and osteocytes) is currently undergoing 
revision. This has resulted from the find- 

ing that some matrix components are 
also synthesized by nonconnective tis- 
sue cells. Of particular interest in this 
regard is fibronectin, which has been 
reported to be distributed on the sar- 
colemma of adult striated muscle. Fur- 
thermore, reports have appeared de- 
scribing changes in fibronectin distri- 
bution and accumulation during the 
differentiation of myoblasts to myo- 
tubes in clonal muscle cell lines in vitro. 
We have carried out similar studies of 
fibronectin biosynthesis secretion and 
extracellular organization during myo- 
genesis in vitro, but we have used pri- 
mary cultures of embryonic muscle cells 
and monoclonal antibody techniques. 
The results have interesting implica- 
tions for muscle cell surface differentia- 



tion, as well as for the production and 
organization of matrix components by 
muscle cells. This study also illustrates 
a variety of techniques that will be im- 
portant in designing similar studies 
concerning other components of extra- 
cellular matrix. 

We generated a monoclonal antibody 
to fibronectin by hybridizing SP 2/0 
myeloma nonsecretor cells with spleen 
cells of a mouse that had been immu- 
nized with chick embryo fibroblast 
fibronectin. We purified the antigen 
by differential extraction and chroma- 
tography on an affinity column of col- 
lagen. The initial screening for positive 
hybridoma clones involved antibody 
binding to live muscle cultures, using 
iodinated or fluoresceinated second an- 
tibody specific for mouse immuno- 
globulin. Cloning in soft agar and fur- 
ther assays led to the selection of a 
single IgG-producing clone. The IgG 
produced by this clone bound to, and, 
in the presence of goat anti-mouse IgG 
antibody, precipitated (1) cell surface 
derived fibronectin, (2) secreted fibro- 
nectin, and (3) cold insoluble globulin (a 
circulatory form of fibronectin). The 
antibody proved to be specific for chick 
fibronectin and did not react with 
fibronectins from horse serum or calf 
serum. This factor has had important 
experimental consequences, for it has 
allowed us to analyze the production 
and distribution of endogenous fibro- 
nectin in both chick muscle and fibro- 
blast cultures, even though exogenous 
fibronectin was present in the serum 
component of the culture medium. (Our 
chick embryo extract did not contain a 
detectable quantity of fibronectin.) 

Because both fibroblasts and myo- 
genic cells synthesize fibronectin, we 
used a method described in Year Book 
79 to generate essentially pure myotube 
cultures. A monoclonal antibody which 
binds to fibroblasts and myoblasts was 
used together with complement to lyse 
all but differentiated myotubes in the 
cultures. We then used these treated 
cultures to examine the kinetics of bio- 

synthesis and secretion of fibronectin. 
Figure 12 illustrates the appearance of 
labeled secretory fibronectin in the cul- 
ture medium as a function of time after 
the introduction of 35 S-methionine. Se- 
cretion of soluble fibronectin can be de- 
tected about 2-3 hours after addition 
of label. During a chase with unlabeled 
methionine or after inhibition of pro- 
tein synthesis, fibronectin secretion con- 
tinues for about two hours. In these 
aspects the kinetics of fibronectin pro- 
duction resemble those described for 
AChE biosynthesis and secretion. In 
fact, analysis of labeled secretory pro- 
teins of "pure" skeletal muscle cultures 
by SDS gel electrophoresis suggests 
that very similar kinetics of synthesis 
and secretion pertain to all the major 
secretory proteins. 

Fibronectin biosynthesis and secre- 
tion have also been studied in chick- 
embryo fibroblasts. In these cells, 
which should be metabolically similar 
to the fibroblasts present in ordinary 
primary muscle cultures, the intracel- 
lular transport time is only about 30 

2 3 4 


Fig. 12. Kinetics of secretion of 35 S-methio- 
nine-labeled fibronectin into the culture medium 
during continuous labeling of chick myotube cul- 
tures (closed circles) and tertiary chick embryo 
fibroblast cultures (open circles) with 35 S-meth- 
ionine. Note lag before acceleration of secretion 
rate to steady level. The time lag is relatively 
brief for fibroblast fibronectin secretion (30 min) 
compared with that observed for pure myotube 
cultures (2-3 hours). 



minutes (as opposed to 2-3 hours for 
myotubes) (Fig. 12). Furthermore, 
these fibroblasts appear to make and 
secrete fibronectin at a rate about five 
times greater per nucleus than do myo- 
tubes. Thus it was imperative to re- 
move fibroblast contamination in order 
to measure reliably myotube fibronec- 
tin secretion. 

An interesting outcome of this study 
was that only about 15% of secretory 
fibronectin is actually released in solu- 
ble form from the myotubes. The re- 
maining 85% of the fibronectin synthe- 
sized in these muscle cultures stays 
bound either to the cells or to the culture 
dish. This fibronectin is trypsin sen- 
sitive, and its accumulation can there- 
fore be measured during labeling with 
35 S-methionine. As in the case of in- 
tegral membrane proteins, fibronectin 
begins to appear at the cell surface 2-3 
hours after the beginning of the label- 
ing period. Thereafter its accumulation 
is approximately linear with time. 

The accumulation of total cell- and 
dish-bound fibronectin has been stud- 
ied as a function of time in cell culture, 
with the aid of our monoclonal chick- 
specific anti-fibronectin antibody di- 
rectly labeled with rhodamine or fluo- 
rescein (Fig. 13). 

These experiments reveal that there 
is a slow buildup of fibronectin as- 
sociated with the collagenous sub- 
stratum of the culture dish. An in- 
tensely staining fibrillar matrix is 
often associated with solitary myo- 
blasts, especially at the ends of the 
bipolar cells where attachment to the 
substratum occurs. As the myogenic 
cells align prior to fusion, this matrix 
becomes more prominent; immediately 
prior to fusion, punctate accumula- 
tions of fibronectin occur at sites of cell 
contact. There is little fibronectin as- 
sociated with the surface of newly 
formed myotubes. As they mature, 
however, deposits of fibronectin ac- 
cumulate under the myotubes, and 
fibrils containing fibronectin extend 
over their upper surface. 

An additional observation of interest 
is that intracellular pools of newly syn- 
thesized fibronectin can be visualized 
after immunofluorescent labeling of 
permeabilized fibroblasts or myotubes. 
These pools appear to be localized within 
perinuclear granules, consistent with 
the possibility that the major fraction 
of intracellular fibronectin is localized 
with the intracellular pool of newly syn- 
thesized ACh receptors in the Golgi ap- 

Extracellular Matrix Organization 

Although the importance of the ex- 
tracellular matrix in embryonic induc- 
tion has long been recognized, it has 
generally been difficult to determine 
the nature and direction of morphoge- 
netic signals. In most cases, it has been 
impossible to demonstrate conclusively 
where a component of the extracellular 
matrix is actually coming from. This 
knowledge is, however, a prerequisite 
for investigating the flow of informa- 
tion between different cell types during 

We would like to use the specificity of 
monoclonal antibodies to determine the 
origin of extracellular matrix proteins 
during muscle morphogenesis. We are 
trying to set up the following ex- 
perimental system: Monoclonal anti- 
bodies will be constructed against 
matrix components of rat fibroblasts 
and of chicken myogenic cells (several 
of which are already available in our 
lab). An antibody that together with 
complement selectively kills prolifera- 
tive chicken but not rat cells (Year Book 
79), allows us to co-cultivate pure 
chicken myotubes with rat fibroblasts. 
With the aid of species-specific mono- 
clonal antibodies it becomes possible to 
investigate which of the components in 
the arising "hybrid" matrix originate 
from fibrogenic, and which from myo- 
genic, cells. This cannot be done easily 
with polyclonal antisera against matrix 
components, which in most known 
cases have to be absorbed in different 



• tf !%* ' 





Fig. 13. Double-label immunofluorescent localization of fibronectin during myogenesis. A, B: 
Phase and fluorescent views of a solitary myoblast 24 hours after plating, stained by indirect im- 
munofluorescence with monoclonal antifibronectin antibody followed by fluorescein-tagged goat 
anti-mouse antibody. C, D: 48 hours after plating, myogenic cells in the process of aligning and fus- 
ing are surrounded by fibrillar network of fibronectin with dense accumulations at sites of close cell- 
cell contact. E, F: In well-developed post-fusion myotube cultures (six days after plating), deposits of 
fibronectin accumulate under the cells, and fibrils containing fibronectin extend from the upper cell 
surface and cell edges to points of substrate contact away from the myotubes. G, H: Ten days after 
plating, dense fibronectin fibrils remain associated with the myotube cell surface. Also note the ac- 
cumulation of a uniform mat of fibronectin bound to the collagenous substratum of the cultures. 



ways to improve specificity. (Across 
different species, common antigenic de- 
terminants of many known matrix com- 
ponents may reflect their evolutionary 
conservative structure. However, since 
the biological activities of, for example, 
fibronectins and collagens are not 
species-specific, it is plausible to as- 
sume that a rat-chick hybrid matrix 
might form a functional unit.) 

With the aid of this experimental 
system, we will try to determine if the 
production of a certain matrix compo- 
nent by, for example, myogenic cells, is 
induced in response to the deposition 
by fibroblasts of another structure 
(and vice versa). The long-term aim of 
such experiments is to determine how 
and where, during muscle morphogene- 
sis, "positional information" is created 
and how it is translated by muscle, con- 
nective tissue, and nerve cells into a 
functional structure. 

In a first approach two antigens were 
chosen: rat fibronectin as a major ma- 
trix component produced by fibro- 
blasts with known actions on myogenic 
cells, and chicken Type V collagen, 
which binds fibronectin and is reported 
to be located in the endomysium sur- 
rounding single muscle fibers (Bailey 
et al, Nature, 278, 67-68, 1979). 

Rat fibronectin was isolated from rat 
serum by affinity chromatography on 
gelatin-Sepharose. The two major 
bands of Mr 235,000 and 220,000 co- 
migrated with the respective horse 
serum fibronectin bands on SDS-gels. 
Four mice were immunized with 200 /xg 
each of this preparation and hybri- 
domas producing anti-fibronectin anti- 
body are being cloned. 

Type V collagen was isolated by pep- 
sin treatment of homogenized 1 6-day - 
old chicken embryos, differential salt 
precipitation, and DE-52 chromatog- 
raphy. The preparation revealed A 
and B collagen chains in a 1:2 ratio on 
SDS gels. A minor contaminant of Mr 
~ 50,000 was still present. For the first 
i.p. injections of four mice (300 fig per 
mouse), we used chicken Type V colla- 

gen kindly provided by Drs. H. and K. 
von der Mark (Munich). Two boosts 
were made after a month, and the fu- 
sion of mouse spleen cells with Sp 2/0 
cells was set up as described elsewhere 
(E. Bayne et al., Year Book 78). Hybri- 
doma supernates were tested in a micro- 
ti ter assay with purified Type V colla- 
gen as a bound antigen and 125 1 -anti- 
mouse Ig as a second antibody. Of 200 
supernates, 45 gave responses of over 
2000 cpm above background. These 
positive supernates were further screened 
by immunofluorescence labeling of 
adult chicken skeletal muscle. Four dif- 
ferent classes of staining patterns were 
found: (1) fibrillar staining of endomy- 
sium and perimysium, (2) fibrillar stain- 
ing of endomysium, (3) basement mem- 
brane staining, and (4) no staining at 
all. All positive hybridoma cultures 
were grown in flasks and frozen; the 
new supernates were tested a third 
time to demonstrate directly antibody 
binding to Type V collagen. For this 
purpose, Type V collagen was labeled 
with 125 I by the chloramine T method. 
Immunoprecipitations were set up by 
incubating labeled collagen with hybri- 
doma supernates followed by addition 
of anti-mouse Ig as a second antibody. 
The precipitates were run on SDS gels, 
which were dried and fluorographed. 
None of the supernates precipitated 
more labeled collagen than the control 
(nonrelated mouse IgG). Certain super- 
nates specifically precipitated the Mr 
-50,000 contaminant of the Type V 
collagen preparation mentioned above; 
these were, with two exceptions, nega- 
tive for immunofluorescence staining. 
We therefore have not yet succeeded in 
unambiguously identifying an anti- 
Type V collagen antibody; despite the 
large number of positives found in mi- 
crotiter and immunofluorescence as- 
says, it remains a possibility that none 
of these antibodies is really directed 
against Type V collagen (which is known 
to be a poor antigen). However, for in- 
vestigations with the rat-chick culture 
system described above, every anti- 



body directed against components of 
endomysium and/or basement mem- 
brane is potentially interesting. 

Skeletal Muscle Insulin Receptors 

Insulin has diverse effects on muscle, 
although its mechanism of action is for 
the most part unknown. Insulin in- 
creases glucose transport across the 
muscle cell membrane, increases the 
glycolytic rate, and stimulates gly- 
cogen, RNA, and protein synthesis in 
muscle. The initial event mediating in- 
sulin action is the binding of insulin to 
a cell surface receptor. Regulation of 
insulin receptor level may be a major 
factor in the modulation of target-cell 
responsiveness in any tissue, including 

We are determining muscle cell sur- 
face insulin receptor concentration and 
affinity under conditions in which in- 
sulin receptor may be regulated. The 
first is developmental. It is apparent 
that insulin plays a supportive role in 
muscle development both in vivo and 
in vitro. We are determining changes in 
insulin receptor number associated 
with differentiation. Previous work 

from this laboratory has shown that 
acetylcholine receptor levels increase 
during differentiation strictly as a re- 
sult of an increased rate of synthesis of 
this cell surface glycoprotein. 

Our present study affords an oppor- 
tunity to compare the mechanism by 
which two different cell surface glyco- 
proteins may be regulated during dif- 
ferentiation. Similar techniques devel- 
oped in this laboratory for directly 
determining rates of synthesis and de- 
gradation of cell surface components 
are being employed in our study of the 
insulin receptor. 

Insulin may be a modulator of its 
own receptor. We are currently investi- 
gating this phenomenon of "down reg- 
ulation" in muscle in order to deter- 
mine the mechanism by which insulin 
receptor level may change in response 
to exposure to its own ligand. 

These experiments will provide infor- 
mation on the mechanism of regulation 
of the specific insulin receptor in mus- 
cle. In addition, we will attempt to 
relate this to other muscle cell surface 
proteins in order that a picture of the 
dynamic state of the entire muscle cell 
surface may emerge. 


R. L. Rotundo and M. R. Emmerling 


The process of development and dif- 
ferentiation in multicellular organisms 
requires not only the correct pattern of 
gene expression in the appropriate cells 
but also the highly ordered spatial and 
temporal distribution of their products 
to the appropriate subcellular loca- 
tions. Nowhere is this complexity more 
beautifully illustrated than in the de- 
velopment of the nervous system, 
where billions of nerve cells form 
highly specific contacts with each 
other and their effector cells. How this 
specific pattern of connectivity is es- 
tablished and maintained by the ner- 

vous system constitutes one of the cen- 
tral problems in neurobiology. 

At all levels of analysis the problem 
of synaptogenesis is one of specificity. 
How is it that a nerve cell recognizes a 
specific target cell as being the appro- 
priate one with which to make and 
establish synaptic contact? Which spe- 
cific molecules are involved in estab- 
lishing and maintaining these contacts 
in order to retain functional connec- 
tivity? How are those molecules tar- 
geted to the appropriate sites of con- 
tact between cells, and at what level is 
this information specified? Finally, we 



might ask, how are the numbers and 
specific destinations of these molecules 
regulated, and what are the respective 
roles of the nerve and the target cell in 
the final outcome? 

The system we have chosen with 
which to begin answering some of 
these questions is the developing 
neuromuscular junction in chick and 
quail. In particular, we are interested 
in the multiple forms of acetylcholines- 
terase (AChE) synthesized by muscle 
cells and the mechanisms of their 
regulation by nerves. Acetylcholines- 
terase is the enzyme that hydrolyzes 
the neurotransmitter acetylcholine re- 
leased by nerves at the neuromuscular 
junction. This enzyme is found in both 
muscle and nerve cells and consists of a 
family of molecular forms distinguished 
by their sedimentation coefficients. The 
most common molecular forms are 
monomers, dimers, and tetramers, plus 
a unique asymmetric form of the en- 
zyme having three tetramers cova- 
lently attached to a three-stranded col- 
lagen-like tail. The tail appears to be 
involved in attaching the asymmetric 
form of the enzyme to the extracellular 
matrix of muscle cells. In addition to 
these oligomeric forms, AChE also ex- 
ists as soluble and membrane-bound 
molecules. The soluble enzyme is se- 
creted by muscle and nerve cells, 
whereas the membrane-bound enzyme 
appears to be an integral membrane 
protein. In muscle cells, these enzyme 
forms are found as part of an in- 
tracellular pool, on the external plasma 
membrane, and they are associated 
with the extracellular matrix which 
envelops these cells. Following inner- 
vation, a large number of several forms 
of AChE molecules accumulate at the 
site of nerve/muscle contact, an area 
constituting less than 0.1% of the total 
muscle surface area. 

The role of the nerve cell, however, 
does not cease once the site of nerve- 
muscle contact has been established. 
We know from the work of many inves- 
tigators that the functional integrity of 

the nerve-muscle contact is important 
for the maintenence of the muscle cell 
as well as the nerve. Pronounced mor- 
phological, physiological, and biochem- 
ical changes occur in both cells when 
their path of communication is inter- 
rupted. One such alteration is the dis- 
appearance of asymmetric AChE from 
the tissue. Another is a dramatic in- 
crease (or decrease, depending upon the 
species) in the levels of AChE in the 
muscle and in its subcellular distribu- 
tion. On the other hand, if functional 
communication between these cells is 
restored by reinnervation, asymmetric 
AChE reappears. Not only does it reap- 
pear, but it accumulates at the ap- 
propriate sites on the cell surface. 

Our current research is centered 
around three general questions. First, 
what are the structural differences be- 
tween the various AChE forms that de- 
termine their tertiary structure and 
their subcellular localization, either 
membrane-bound, secreted, or attached 
to basal lamina? Second, what is the se- 
quence of events involved in the syn- 
thesis and assembly of the several mo- 
lecular forms of the enzyme? Third, are 
the several forms of this enzyme en- 
coded by separate genes, or do they 
arise by differential processing and/or 
post-translational modifications of a 
common precursor? These studies will 
form the basis for our subsequent stud- 
ies on the mechanism(s) of regulation 
of the molecular forms of AChE by 
nerves. Ultimately, what we hope to 
gain is an understanding of how de- 
fined cell-cell interactions during de- 
velopment, in this case the specific in- 
teractions between nerve and muscle 
cells, lead not only to changes in the ex- 
pression of specific genes, but also to 
changes in the synthesis, processing, 
and redistribution of their products. 

In this report, we describe initial 
work on the purification of brain AChE 
and some preliminary studies on the 
enzyme forms isolated from brain and 
muscle tissue. In addition, several 
studies concerning the synthesis of 



AChE forms of quail muscle cultures 
are described. This system is unique in 
that the quail muscle cells synthesize a 
20S asymmetric form of AChE in the 
absence of nerve cells. Muscle cells 
from chicken embryos do not synthe- 
size this form without being inner- 
vated. Finally, preliminary studies 
aimed at understanding the interac- 
tions of the 20 S asymmetric AChE 
with other extracellular matrix com- 
ponents, such as fibronectin, are 

Purification of 

R. L. Rotundo 

Our initial task in this project was to 
obtain purified acetylcholinesterase for 
the production of antibodies as well as 
for studies of the molecules themselves. 
This task was complicated by the fact 
that AChE exists as several molecular 
forms with different properties in most 
tissues and because in all tissues 
AChE constitutes less than 0.01% of 
total protein. In other words, a 10,000- 
fold purification is necessary to obtain 
a homogenous protein. We chose to 
purify the 11.5S tetrameric form of 
AChE from adult chicken brain, since 
this form constitutes about 90% of the 
total enzyme activity in this tissue (see 
Fig. 14). Based on our current knowl- 
edge of AChE properties, the enzyme 
isolated from brain tissue will be suf- 
ficiently similar to that of muscle so 
that cross-reacting antibodies can be 

After synthesizing and testing many 
potential affinity ligands for AChE, we 
succeeded in constructing one that has 
a sufficiently high affinity for the en- 
zyme that AChE molecules will bind to 
it in solution yet can still be detached 
following the addition of a water-solu- 
ble inhibitor. The principle of this chro- 
matographic procedure is to attach a 
ligand, in our case an aromatic quarte- 
nary ammonium compound, to a spacer 

molecule, which in turn is attached to 
derivatized agarose beads. When a so- 
lution containing AChE molecules is 
passed through the column, the enzyme 
is selectively bound. After washing un- 
bound proteins from the column, the 
AChE is removed by eluting with a sol- 
uble inhibitor. This purification proce- 
dure is most effective when preceded by 
ion-exchange chromatography to re- 
move several proteins which tend to bind 
nonspecifically to our column. The effi- 
ciency of this purification procedure is 
shown in Fig. 15. The fractions from 
the affinity column containing AChE 
are further purified by preparative gel 
electrophoresis, and the protein is then 

10 20 

Fraction number 

Fig. 14. Molecular forms of acetylcholines- 
terase in adult chicken brain. Brain tissue was 
homogenized in 5 volumes 10 mM Tris pH 7, 1 
mM EDTA, and 0.5% Triton-100, followed by 
centrifugation at 15,000 X g. A 100-/d aliquot 
was analyzed by velocity centrifugation in a 
5-20% sucrose gradient and assayed for AChE 
activity, as described by Rotundo et al. (J. Biol. 
Chem., 254,4790-4799, 1979). The 11. 5S peak 
contains approximately 90-95% of the total en- 
zyme activity. 




— 130 

pi 100 

«** 68 

mmf 57 

PI 43 
mm 31 




9 * 



Fig. 15. SDS-gel electrophoresis of fractions obtained during AChE purification. (A) Coomassie 
stained gel: (a) Whole brain homogenate, (b) 13,000 X g supernatant, (c) effluent from ion exchange 
column, (d) column wash, (e) elute peak from ion exchange column, (f) protein retained by ion ex- 
change column, (g) effluent from affinity column, (h) eluted peak from affinity column. The outside 
lanes are molecular weight markers. (B) Autoradiograph of 3 H-DFP-labeled AChE. An aliquot of 
AChE from fraction (h) was labeled with 3 H-diisopropylfluorophosphate, which covalently binds to 
the active site of the enzyme. After SDS-gel electrophoresis, the gel was treated with "Enhance" and 
fluorographed. The AChE monomer is the major band at approximately 115,000 daltons molecular 
weight. The lower-molecular-weight bands may be degradation products formed during purification. 

eluted from the gel (the 115,000-dalton 
band in lane h, Fig. 15 A). Alternatively, 
acetylcholinesterase can be reacted 
with 3 H-diisopropylfluorophosphate 
( 3 H-DFP), which binds irreversibly to 
the active site of the enzyme. The la- 
beled enzyme can then be analyzed by 
gel electrophoresis and autoradio- 
graphed (Fig. 15B). This procedure is 
useful for identifying both the intact 
active-site-containing protein and break- 
down products of the enzyme. Follow- 
ing preparative gel electrophoresis, the 
enzyme is used for the immunization of 

rabbits and mice or in screening proce- 
dures during production of monoclonal 

Comparison of AChE Molecules 
Isolated from Brain and Muscle 

Neurons and muscle cells both con- 
tain acetylcholinesterase molecules 
having sedimentation coefficients of 
approximately 7S and 11. 5S. Using 
adult chicken brain as a source of 
neuronal AChE and cultured chick em- 
bryo muscle cells as our source of mus- 



cle AChE, we find that in both cases 
about 90% of the enzyme is membrane 
bound. However, analysis of the 
3 H-DFP- labeled AChE from brain and 
muscle tissue by SDS-gel electropho- 
resis indicates that their subunits may 
be quite different (Fig. 16). Whereas the 
AChE from neuronal sources has an ap- 
parent monomer molecular weight of 
115,000 daltons (Fig. 15), the enzyme 
isolated from muscle appears as two 
bands of apparent molecular weights 
86,000 and 92,000 daltons when run 
under identical conditions. Although 
all possible artifacts have not yet been 
eliminated, we do not find breakdown 
products in the 85,000-90,000 dalton 
range when analyzing the brain AChE. 
These observations raise the interest- 








- 8 

1 \ 

■ 7«r- 



■ 6 o 

• 5 — 
4 * 

CL 100 



O 80 

N -^ 



- 3 

- 2 






1 1 V*tlMfV*l»lllll 1 1 1 

10 20 30 40 50 60 

Slice number 

Fig. 16. SDS-gel electrophoresis of 3 H-DFP- 
labeled AChE isolated from cultured chick em- 
bryo muscle cells. Partially purified AChE was 
labeled with 3 H-DFP and run on an SDS- 
gel identical to that used in Fig. 15B. The lane 
containing AChE was excised, sliced, and the ra- 
dioactivity determined by liquid scintillation 
counting. An adjacent lane was treated with 
"Enhance" and fluorographed. The AChE pres- 
ent in cultured muscle cells appears as two labeled 
bands with apparent molecular weights of about 
86,000 and 92,000 daltons. 

ing possibility that the neuronal and 
muscle AChE forms are substantially 
different from each other. The nature 
of these differences is currently under 

Studies on Molecular Forms 

of Acetylcholinesterase in Quail 

Muscle Cell Cultures 

M. R. Emmerling 

Velocity sedimentation analysis of 
AChE from quail muscle cell cultures 
revealed the presence of at least four 
molecular forms of AChE. These forms 
can be distinguished by their sedimen- 
tation coefficients, 20.7S, 11.8S, 7.3S 
and 5.4S. Of the four, only the 20.7S 
form is sensitive to collagenase diges- 
tion and it self-aggregates under condi- 
tions of low ionic strength. This species 
of AChE is analagous to the collagen- 
tailed forms of AChE found in the 
electric eel and other vertebrates. The 
presence of the collagen-tailed form in 
muscle cell cultures provides a valuable 
opportunity to study the assembly and 
fate of this macromolecule in a defined 
cellular system. Current questions un- 
der investigation include the precursor- 
product relationships between the 
20. 7 S and smaller forms of AChE, the 
site of assembly of the 20. 7 S enzyme, 
and the mechanism by which the 20.7S 
form becomes an extracellular enzyme. 
The collagen-tailed form of AChE is of 
particular interest to neurobiologists, 
as it appears to be a neurally regulated 
component of basement membrane at 
motor endplates. Thus, this form can 
be used as a representative molecule for 
studying those processes involved in 
the organization of extracellular ma- 
trix at developing neuromuscular junc- 

Cellular Distribution of AChE Forms 

Because the 20.7S collagen-tailed 
AChE appears to be an extracellular 
enzyme at neuromuscular junctions in 



vivo, experiments were performed to 
determine whether this form of AChE 
was also present outside muscle cells in 
vitro. To perform this experiment, a 
method was devised to distinguish ex- 
tracellular from intracellular AChE. 
The method uses pharmacological in- 
hibitors of AChE that do not easily 
pass through plasma membranes 
(Mclssac et al, J. Pharmacol. Exp. 
Ther., 126, 9-20, 1959; Rotundo et al, 
Cell, 22, 583-594, 1980). In my own 
work, I took advantage of the fact that 
ectothiophate iodide (ECHO), an ir- 
reversible, organophosphate inhibitor 
of AChE, is positively charged and 
under appropriate conditions does not 
penetrate cultured muscle cells. Thus, 
at 4 °C10- 5 M, ECHO can be used to in- 
hibit selectively extracellular AChE. 

Analysis of the molecular forms of 
AChE in muscle cell cultures treated 
with ECHO was performed to deter- 
mine which forms were extracellular. 

Since ECHO inhibits only extracellular 
AChE, the difference in enzyme activ- 
ity between untreated and ECHO- 
treated cultures represents the amount 
of extracellular AChE activity, in this 
case 36% of total AChE in control 
cultures. As can be seen in Fig. 17, 
ECHO treatment resulted in a decrease 
in activity of all molecular forms of 
AChE, demonstrating that in these 
cultures all AChE forms are present 
both inside and outside the cells. Com- 
puting the difference between the area 
under the AChE peaks from control 
and ECHO-treated cultures, we esti- 
mate that approximately equal amounts 
of the forms 20.7S, 11. 8S, and 7.3S + 
5.4S (these forms are combined, as 
they do not separate well on gradients) 
are present extracellularly. Although 
tentative, this result supports the hy- 
pothesis that the collagen-tailed form 
is assembled intracellularly and trans- 
ported to the cell surface, as shown for 






Fig. 17. Comparison of AChE forms in control (open circles) and ECHO-treated (closed circles) 
quail muscle cell cultures. ECHO-treated cultures were incubated in defined medium containing 10~ 5 
M ECHO for 60 min at 4°C. Control cultures were treated in the same manner but without ECHO. 
AChE in cultures was extracted with 20 mM borate buffer (pH 8.9) containing 1.0 M NaCl and 0.5% 
Triton X-100. AChE samples were analyzed by velocity sedimentation on 5-20% linear sucrose gra- 
dients. Fractions of gradients were assayed for AChE activity. AChE activity is expressed in ar- 
bitrary units. 



smaller forms of AChE (Rotundo et al. 
Cell, 22, 583-594, 1980). Future ex- 
periments will focus on the site of 20. 7S 
AChE assembly in quail muscle cul- 
tures and will determine whether this 
species is transported to the cell sur- 
face in the same manner as smaller 
AChE forms. 

Effects of Defined Medium on 
AChE Molecular Forms 

It has been observed in other cell 
culture systems that serum in culture 
medium can affect the amount of en- 
zymes in cultured cells (Levin et al, 
Cell, 22, 701, 1980). To examine whether 
undefined components in culture me- 
dium can affect the amount or molecu- 
lar forms of AChE in quail muscle cell 
cultures, parallel sets of cultures were 
incubated in either complete medium 
or defined medium (without horse 
serum or embryo extract). Formulation 
of the defined medium is described by 
Konigsberg {Methods Enzymol, 53, 

511, 1979). After nine hours of incuba- 
tion in their respective media, cultures 
were analyzed for AChE activity and 
molecular forms. There was no differ- 
ence in the specific activity of AChE 
under these culture conditions. How- 
ever, the cultures did differ in the 
amount of the different AChE molecu- 
lar forms present. 

Cultures incubated in defined medium, 
as compared to those incubated in com- 
plete medium, exhibited a three-fold in- 
crease in the activity of the 20. 7 S en- 
zyme peak and an equivalent decrease 
in the 7.3S and 5.4S forms (Fig. 18). In 
addition, it appears that the second 
largest AChE form from cultures incu- 
bated in defined medium sediments 
slower than the corresponding 11. 8S 
form from cultures incubated in com- 
plete medium. These changes can be re- 
versed by returning the defined-medium- 
treated cultures to complete medium 
for nine hours. Thus, it appears that 
some component in horse serum or em- 
bryo extract can modulate the levels of 






-i r 

10 20 


Fig. 18. Effect of defined medium on AChE forms in quail muscle cell cultures. Cultures were in- 
cubated for 9 hours in either complete (open circles) or defined (closed circles) medium. After 9 hours, 
AChE was extracted from cultures and run on sucrose gradients, as described in Fig. 17. AChE ac- 
tivity in each fraction is presented as the percent of total AChE activity in the gradients. 



AChE forms without changing the spe- 
cific activity of AChE in quail muscle 
cell cultures. 

Preliminary evidence suggests that 
the horse serum in complete medium 
may inhibit the appearance of 20.7S 
AChE in quail muscle cell cultures by 
affecting its synthesis and/or its degra- 
dation. Future studies on the synthesis 
of the 20.7S will include the use of de- 
fined medium to control for effects of 
the undefined components of complete 

Binding of Collagen-tailed AChE to 
Fibronectin on Microtiter Dishes 

Fibronectin is a cell surface glycopro- 
tein which can interact both covalently 
and noncovalently with collagen. In 
previous work (Emmerling et al, Bio- 
chemistry, 20, 3242-3247, 1981), it was 
shown that only the collagen-tailed 
forms of AChE could interact with fi- 
bronectin. Using velocity sedimenta- 
tion in sucrose gradients, it was possi- 
ble to show that fibronectin could 
change the sedimentation profile of 
collagen-tailed AChE. However, be- 
cause of the cumbersome nature of this 
analytical method, it was not possible 
to characterize completely the interac- 
tions between these two molecules. To 
continue the experiments, a rapid and 

simple technique was devised to study 
the interaction between collagen-tailed 
AChE and fibronectin. The method 
stems from work of Engvall et al. (J. 
Exp. Med., 147, 1584, 1978), who 
showed that collagen could bind to the 
fibronectin attached to polystyrene 
microtiter plates. Using this method, it 
was possible to bind the 20. 7S collagen- 
tailed AChE to fibronectin-coated wells 
(the slower sedimenting forms do not 
bind to fibronectin). As had been ob- 
served in earlier studies, binding be- 
tween 20.7 S AChE and fibronectin was 
inhibitable by 50 mM spermidine (an 
inhibitor of fibronectin-collagen inter- 
actions) and 0.2 M NaCl. However, once 
collagen-tailed AChE was bound to fi- 
bronectin, the AChE form could only 
be partially removed by treatment with 
100 mM spermidine and 1.0 M NaCl, 
suggesting that, once attached, AChE 
binds tightly to fibronectin. 

These results demonstrate that colla- 
gen-tailed AChE interacts with fibronec- 
tin bound to microtiter wells. The inter- 
action appears to be similar to that 
already described for AChE and fibro- 
nectin. The simplicity of the method 
makes it useful for the further charac- 
terization of fibronectin-AChE interac- 
tions. These studies will be important 
for understanding the mechanism(s) of 
association of the 20.7S AChE with the 
cell surface in quail muscle cultures. 


D. Hoekstra, K. J. Longmuir, O. C. Martin, J. W. Nichols, and R. E. Pagano 

This laboratory has a long-standing 
interest in the organization, dynamics, 
and metabolism of lipid molecules in 
the membranes of eukaryotic cells. Our 
principal aim is to dissect the patterns 
of translocation of the major lipid com- 
ponents between, for example, the cell 
surface and internal membranes, and 
between various internal membrane 
pools and their sites of de novo syn- 

thesis. Ultimately we hope to be able to 
correlate this data with the ever- 
growing body of information on mem- 
brane-protein processing and with 
physiologically important events such 
as hormonal and electrical stimulation 
of cells. 

Due by and large to a battery of tech- 
niques which allow one to selectively 
label a particular membrane protein in 



a particular cellular compartment so 
that its fate can be followed with time, 
considerable insights have been gained 
into the mechanism(s) by which mem- 
brane glycoproteins are inserted into 
the cell surface and are subsequently 
metabolized. Such methodology in- 
cludes radio-iodination of cell surface 
proteins, immunochemical techniques, 
and the tagging of specific proteins 
with labeled toxins or hormones— tech- 
niques which are not available for the 
study of cell membrane lipids because 
of their chemical nature. Although a 
considerable body of data exists on the 
"bulk," or "whole-cell," metabolism of 
cellular lipids, little if anything is 
known about the pathways of translo- 
cation of membrane lipids between var- 
ious compartments within cells, and 
about how these pathways correlate 
with lipid metabolism. For example, it 
remains to be demonstrated whether 
newly synthesized lipid molecules 
make their way to the cell surface via 
the now generally accepted pathway 
for membrane protein synthesis which 
requires participation of the Golgi ap- 
paratus of the cell, or whether alter- 
native pathways such as exchange or 
net transfer of lipids, known to occur in 
vitro, also play an important role. Simi- 
larly, essentially no information is at 
hand on the residence time for different 
phospholipid species in the plasma 
membrane of a cell and their route of in- 
ternalization upon leaving the cell sur- 

In order to address these problems, 
earlier studies from this laboratory 
focused on developing technology for 
introducing appropriately tagged (i.e., 
radioactive, fluorescent, or haptenated) 
phospholipid molecules (or analogs) into 
cells via phospholipid vesicles so that 
their subsequent processing could be 
studied (see Year Book 78, 79). Under 
appropriate conditions, for example, 
significant quantities of fluorescent 
analogs for the end-products of lipid 
metabolism (phosphatidylcholine or 
phosphatidylethanolamine) can be in- 

troduced into the cell surface without 
labeling internal lipid pools. The fate of 
such molecules as they move from the 
cell surface to internal compartments 
can then be followed biochemically and 
by fluorescence microscopy as a func- 
tion of time. In this Report, we present 
the progress we have made toward the 
various cell biological problems outlined 
above. We also describe supporting 
studies on the basic physical properties 
of a number of phospholipid analogs 
used in this work. 

Kinetics of Phospholipid Transfer 
between Vesicles 

J. W. Nichols 

There is increasing evidence that bio- 
logically important amphiphilic mole- 
cules, such as phospholipids, fatty 
acids, and cholesterol, can be passively 
transferred between various combina- 
tions of vesicles, lipoproteins, and cellu- 
lar membranes, as soluble monomers dif- 
fusing through the aqueous medium. We 
reported in last year's Report the use of 
a fluorescently labeled phospholipid, 
zole-4-yl)amino]caproyl] phosphatidyl- 
choline (C 12 -NBD-PC), to investigate the 
mechanism of amphiphile transfer be- 
tween two populations of vesicles. The 
C 12 -NBD-PC transfer was monitored as 
it moved from donor vesicles containing 
self-quenching concentrations of this 
fluorescent molecule into unlabeled ac- 
ceptor vesicles. Kinetic measurements of 
initial transfer rates indicated that 
transfer of this molecule occurs via the 
diffusion of soluble monomers. During 
this year, we have further studied this 
soluble monomer diffusion process using 
resonance energy transfer to monitor the 
movement of fluorescent phospholipids 
between vesicles. This technique differs 
from the first in that it allows measure- 
ment of the transfer of probe molecules 
which comprise only a small fraction of 
the total vesicle phospholipid. By 
measuring the transfer of a small frac- 



tion of the total phospholipid, we allow 
complete equilibration of the fluorescent 
phospholipid molecules to occur without 
significant rearrangement of the donor 
or acceptor vesicles. Consequently, both 
rate and equilibrium measurements are 
possible. By combining the information 
gained from both types of measure- 
ments, we can determine the effect of the 
amphiphile structure and vesicle com- 
position on both the rate at which the 
amphiphile enters and leaves the vesicle 
bilayer, and the affinity for vesicles of a 
given phospholipid composition. 

The relative ease with which transfer 
measurements can be made with these 
fluorescent probes makes them ideally 
suited for studies of general principles 
governing the transfer of amphiphiles 
via the diffusion of soluble monomers. 
From these studies we have concluded 
that (1) the rate at which an amphiphile 
enters or leaves a bilayer depends on 
the structure of both its polar and non- 
polar regions, (2) the on-rate, off-rate, 
and affinity of a given amphiphile will 
differ depending on the lipid composi- 
tion of the vesicle, and (3) for two vesicle 
populations having different off-rates 
for a given amphiphile, the half-time for 
equilibration will vary depending on 
the concentration ratio. Our hope is 
that these principles will prove gener- 
ally useful in understanding passive 
diffusion of biologically relevant am- 
phiphiles between membranes. 

Studies of Membrane Fusion in 
Vesicle-Vesicle and Vesicle- 
Cell Systems 

D. Hoekstra 

In Year Book 79, we described a 
novel assay that employed resonance 
energy transfer to examine Ca 2+ -in- 
duced fusion of phosphatidylserine 
(PS)-containing lipid vesicles. Using 
this method, fusion can be monitored 
continuously by following the increase 
in transfer efficiency between a fluores- 
cent donor (Af-NBD-PE) and acceptor 

(JV-Rh-PE) molecule when one popula- 
tion of PS vesicles containing a small 
amount of iV-NBD-PE is fused with a 
second population containing a small 
amount of AT-Rh-PE. Although the role 
of Ca 2+ in fusion phenomena has been 
well documented, the mechanism by 
which Ca 2+ acts to promote fusion is 
still unclear. In fact, several mecha- 
nisms have been proposed to explain 
the fusion process in molecular terms, 
one of which involves the occurrence of 
phase separations in the bilayer upon 
addition of Ca 2+ . This may lead to 
changes in molecular packing and the 
creation of phase boundaries. It has 
been suggested that phase boundaries, 
in turn, may represent a thermodynam- 
ically favorable site for the initiation of 
membrane fusion. Proper experimental 
proof of this hypothesis has been ham- 
pered by the lack of a suitable assay to 
measure the kinetics of phase separa- 
tion. However, we have now been able 
to develop a method which monitors 
this process continuously during the 
time course of vesicle-vesicle fusion. 
The method is based on the use of a 
small amount of a fluorescent phos- 
pholipid analog (C 6 -NBD-PC or 
AT-NBD-PE) incorporated into the PS 
vesicle bilayer which, during Ca 2+ -in- 
duced phase separation, is concen- 
trated in restricted domains of the 
membrane and becomes self-quenched, 
an event presumably due to intermolec- 
ular quenching. In Fig. 19, the results 
of an experiment are shown wherein 
PS/Af-NBD-PE vesicles (95:5) were 
fused with PS/AT-Rh-PE vesicles (95:5) 
in the presence of Ca 2+ (solid curves). 
The dashed curves indicate the kinetics 
of Af-NBD-PE quenching as a result of 
phase separation when PS/Af-NBD-PE 
vesicles alone are incubated with cor- 
responding Ca 2+ concentrations. The 
results show that the fusion process 
proceeds much faster than the process 
of phase separation. For example, in 
the presence of >4 mM Ca 2+ , the fu- 
sion process proceeds at a rate of tu 2 
— 1 min for the phase separation pro- 





\ \ 

\ v 

\ \ 
\ \ 
\ \ 



\ \ 
N \ 

V ^ Jj~--^ J) 



^ >^ 


i i i i i 

2 4 6 8 10 12 

Time (min) 

Fig. 19. Kinetics of fusion and phase separa- 
tion of PS vesicles as a function of Ca 2+ concen- 
tration. PSA/V-NBD-PE alone or PS/7V-NBD-PE 
and PS/N-Rh-PE combined vesicles were in- 
cubated in the presence of Ca 2+ . NBD fluores- 
cence quenching, resulting from self-quenching 
in the phase separation process (dashed curves) 
or resonance energy transfer during fusion (solid 
lines), was monitored as a function of time. Ca 2+ 
concentrations were (a) 2, (b) 4, and (c) 8 mM. 
(NBD quenching resulting from phase separa- 
tion, in contrast to fusion, is reversible upon 
addition of EDTA and thus enables us to deter- 
mine the contribution of NBD quenching due to 
phase separation during vesicle fusion. Relief of 
NBD fluorescence was not observed when EDTA 
was added during or after vesicle fusion, indicat- 
ing that the quenching was exclusively due to 
resonance energy transfer in the latter process.) 

cess. These results suggest that phase 
separation is not directly involved in 
the initiation of vesicle fusion. 

Further evidence supporting this in- 
terpretation was provided by experi- 
ments showing that (1) Mg 2+ induces 
substantial fusion of vesicles (albeit 
at higher cation concentrations than 
Ca 2+ ) without the occurrence of phase 
separation, and (2) subthreshold con- 
centrations of both cations can induce 
extensive vesicle fusion without phase 
separation when the vesicles are in- 

cubated in media containing the dehy- 
drating agent poly(ethylene)glycol 
(PEG). From these results we conclude 
that phase separation does not have to 
be complete for fusion to occur. How- 
ever, the degree of interbilayer dehy- 
dration and/or the creation of "point- 
defects" (as a result of metal-ion-PS 
complex formation) are major factors 
involved in the initiation of fusion. 

Topology of Membrane Fusion 

An alternative approach for gaining 
insight into the molecular events of 
membrane fusion involves the study of 
lipid topology following the fusion pro- 
cess. If membrane topology is main- 
tained, lipids initially present in the 
outer leaflet of the membrane bilayer 
should remain localized in the external 
leaflet after fusion. We have studied 
membrane topology following vesicle- 
vesicle and cell-cell fusion. 

Vesicle-vesicle fusion. Small uni- 
lamellar vesicles (SUV) consisting of 
PS and PE at different molar ratios and 
containing a trace amount of 3 H-PE 
were fused in the presence of Ca 2+ . The 
orientation of PE was determined 
before and after fusion by using an im- 
permeable amino-reactive agent (TNBS) 
that would react only with externally, 
but not internally, exposed amino 
groups at the vesicle surface. The 
results showed that PE redistribution 
did not occur during fusion of these 
vesicles. In contrast, the accessibility 
of PE for TNBS increased dramatically 
(from 10% to 70%) when SUV were 
fused with multilamellar vesicles 
(MLV) consisting of PS and PE. A 
similar phenomenon was observed 
when SUV were "prefused" to form 
large unilamellar vesicles (LUV), which 
were then fused with subsequently 
added SUV. However, redistribution 
was not observed when MLV or LUV 
alone were incubated in the presence of 
Ca 2 +. While we do not yet have a clear 
explanation for these observations, the 
results suggest that redistribution of 



PE is not a necessary consequence of 
fusion per se (as evidenced by SUV- 
SUV fusion), but may be related to 
both fusion and bilayer curvature 
(SUV-MLV and SUV-LUV fusion). 

Cell-cell fusion. The topology of mem- 
brane fusion was also examined during 
cell-cell fusion induced by Sendai virus 
or PEG. The fluorescent phospholipid 
analog C 6 -NBD-PC was inserted into 
the outer leaflet of the plasma mem- 
brane of V79 cells {Year Book 78, 
33-37) by incubating cells at 2°C with 
DOPC/C 6 -NBD-PC vesicles. Redistri- 
bution of the probe in washed, vesicle- 
treated cells could then be detected by 
determining the exchangeable pool be- 
fore and after cell-cell fusion by "ex- 
tracting" C 6 -NBD-PC from the outer 
leaflet with excess unlabeled DOPC 
vesicles (Year Book 79). As shown in 
Fig. 20, decreasing amounts of the flu- 
orophore are recovered from the cell 
surface when the fusion index in- 
creases. We estimate that approxi- 
mately 50% of the probe becomes inter- 
nalized during Sendai virus-induced 
cell fusion. Interestingly, this finding 
seems to support a recent hypothesis, 
based on EM observations, that mem- 
brane fragments may become internal- 
ized during virus-induced fusion. 

Our lab is also exploring membrane 
lipid topology during PEG-induced 
cell-cell fusion; preliminary results in- 
dicate that, in contrast to the results 
obtained with Sendai virus, no substan- 
tial reorganization occurs. This result 
seems to be consistent with our obser- 
vations on the topology during SUV- 
SUV fusion and further suggests that 
C 6 -NBD-PC redistribution during vi- 
rus-induced fusion is not directly re- 
lated to the actual fusion event. 

Internalization of Plasma 
Membrane Phospholipids 

R. E. Pagano and O. C. Martin 

We have previously shown that it 
is possible to integrate exogenously 

20 40 60 80 100 


Fig. 20. Redistribution of C 6 -NBD-PC follow- 
ing Sendai-virus-induced cell-cell fusion. V79 
hamster fibroblasts were fused with increasing 
amounts of virus (to increase the percentage of 
cell fusion) after insertion of C 6 -NBD-PC into 
the outer leaflet of the plasma membrane. The 
percentage of fluorescent probe remaining in the 
cell surface as a function of the fusion index was 
determined as described in the text. 

supplied, isotopically labeled phos- 
phatidylcholine (PC), phosphatidyleth- 
anolamine (PE), or their fluorescent 
NBD-analogs (NBD-PC or NBD-PE) 
into the plasma membrane of several 
cell types by incubating the cells with 
lipid vesicles containing these mole- 
cules at 2°C (Year Book 78, 79). As long 
as the washed, vesicle-treated cells are 
maintained at low temperature, both 
the isotopically labeled and flu- 
orescently tagged lipid molecules re- 
main restricted to the plasma mem- 
brane bilayer (see Fig. 22a), where they 
are oriented toward the external bath- 
ing medium. 

In the present study, we examined 
the fate of fluorescent NBD-PC or 
NBD-PE molecules introduced into the 
plasma membrane of Chinese hamster 
fibroblasts that were warmed to 37 °C 
for varying periods of time. Upon 
warming, plasma membrane fluores- 
cence decreased with the concomitant 
appearance of bright areas, or "specs," 



of fluorescence within the cell's cyto- 
plasm (Fig. 21). The process of plasma 
membrane lipid internalization had a 
half-time of approximately 1.5-2 hours 
and could be blocked by the metabolic 
inhibitors azide and deoxyglucose. Bio- 
chemical analysis of cells treated with 
NBD-PC vesicles for 60 mm at 2 °C and 
subsequently warmed to 37 °C for 90 
min showed no breakdown of the fluor- 
escent phosphatidylcholine analog. 
Furthermore, fluorescence microscopy 
of the cells showed that, over the time 
course of the experiment, none of the 
internalized fluorescent lipids was ap- 
parently reinserted into the plasma 
membrane pool. 

In parallel experiments, the internal- 
ization of a fluorescent lectin, rhoda- 
mine-conjugated Lens Culinaris Agglu- 
tinin, initially bound to cell surface 
glycoproteins at 2°C, was also followed 
after warming the cells to 37 °C. The 
lectin appeared to internalize much 
more rapidly than the fluorescent lipid, 
and it co-localized with the internalized 
fluorescent phospholipid. This result 

demonstrates that cell surface lipids 
are interiorized via an endocytic mecha- 
nism to a particular subcellular com- 
partment, possibly the lysosomal ap- 
paratus, without equilibration with 
other intracellular compartments. Our 
findings further suggest that the path- 
ways for insertion and removal of cell 
surface lipids may be distinct from one 

Cellular Distribution and 

Metabolism of Fluorescent 


K. J. Longmuir, O. C. Martin, D. K. Struck, and 
R. E. Pagano 

An important goal of our laboratory 
is to define the relationship between 
the intracellular distribution and the 
metabolism of phospholipids in mam- 
malian cells. Our research strategy is to 
introduce into cells fluorescent phos- 
pholipids observable by fluorescence 
microscopy. Then, after a defined in- 
cubation period, we extract the lipid 

Fig. 21. Fluorescence photomicrograph of Chinese hamster V79 fibroblasts treated with NBD- 
PC-containing vesicles for 60 min at 2°C and subsequently warmed to 37°C for 90 min. 



fraction and determine how the lipid 
probe has been biochemically modified. 
Hence we can correlate information 
about the location of lipid in intact cells, 
as determined by fluorescence micros- 
copy, with biochemical events in lipid 

We use the fluorophore NBD (N-4- 
nitrobenzo-2-oxa-l,3-diazole) for visual- 
izing the cellular distribution of phos- 
pholipids. This fluorophore is cova- 
lently attached to the acyl chain of a 
variety of phospholipids prepared in 
our laboratory. We introduce these flu- 
orescent lipid probes into mammalian 
cells first by forming unilamellar ves- 
icles containing both the probe and car- 
rier phospholipids. These vesicles are 
then incubated either with monolayers 
or with suspensions of mammalian 
cells in vitro. Currently, most of our 
studies are carried out using the V79 
Chinese hamster lung fibroblast cell 

Our laboratory has reported previ- 
ously on the incorporation of NBD- 
labeled phosphatidylcholine (NBD-PC) 
into intact cells (Year Book 78). When 
NBD-containing vesicles and cells are 
incubated at 2°C for one hour, the 
NBD-PC is incorporated into the 
plasma membrane. Little if any flu- 
orescent phospholipid is observed in in- 
tracellular compartments. However, we 
have now discovered that the NBD 
analog of phosphatidic acid (NBD-PA), 
an intermediate in lipid metabolism, is 
both taken up and rapidly internalized, 
even at 2°C. This remarkable difference 
in the cellular distribution of NBD-PC 
and NBD-PA is seen in Fig. 22. In Fig. 
22a, the fluorescent lipid NBD-PC pre- 
viously shown to be restricted to the 
plasma membrane bilayer (see Year 
Book 78) results in peripheral ring 
fluorescence. By contrast, the cell- 
associated NBD-PA appears to be to- 
tally intracellular (Fig. 22b). Labeling 
of the nuclear membrane is particularly 
prominent. We note that this pattern of 
labeling is strikingly similar to pub- 
lished micrographs of fluorescent anti- 

Fig. 22. Incorporation of NBD-labeled phos- 
pholipids into V79 Chinese hamster lung fibro- 
blasts. Cells were incubated with vesicles con- 
taining 95% dioleoylphosphatidylcholine and 
5% NBD-labeled phospholipid for 1 h at 2°C. (a) 
Cells incubated with NBD-PC. (b) Cells in- 
cubated with NBD-PA. 

bodies directed against components of 
the rough endoplasmic reticulum of 
mammalian cells. 

It is clear that the NBD-PA is prefer- 
entially incorporated into the cell via 
lipid transfer, the mechanism by which 
other NBD-lipids are taken up by cells 
under similar incubation conditions. 
This mechanism was confirmed by in- 
cubating cells with vesicles containing 
both acyl-chain-labeled NBD-PA and 
head-group-labeled rhodamine PE. 
When vesicles containing both these 
probes were incubated with V79 cells, 



we found a disproportionate cellular 
uptake (20:1) of NBD-PA compared to 
rhodamine PE. This result means that 
the uptake of intact vesicles is not a 
significant event in the incorporation 
of NBD-PA. Rather, the NBD-PA is 
preferentially incorporated by the 
transfer of individual NBD-PA mole- 
cules into the cell. 

When incorporated, the NBD-PA is 
metabolized along biosynthetic routes 
characteristic of mammalian cells (Fig. 
23, lanes 2 and 4). Even at 2°C, much of 
the cell-associated NBD-PA is con- 
verted to NBD-diglyceride. A small 
percentage of the probe is further con- 
verted to NBD-PC and NBD-triglycer- 
ide. If vesicle-cell incubations are car- 
ried out at 37 °C, a greater percentage of 
the NBD-PA is converted to NBD-PC 
and NBD-diglyceride. By contrast, un- 
der the conditions of our experiments, 

exogenously supplied NBD-PC was not 
metabolized, either at 2°C or 37 °C (Fig. 
23, lanes 3 and 5). 

Our observation that NBD-PA is 
converted to end products of lipid me- 
tabolism such as PC and triglyceride is 
significant for several reasons. First, 
the data indicate that the enzymes of 
lipid metabolism can catalyze reactions 
involving phospholipids even when the 
NBD fluorophore is present on the 
molecule. Indeed, we found that the 
metabolism by V79 cells of NBD-PA 
and radiolabeled PA was qualitatively 
the same. Second, the fact that NBD- 
PA is converted by biosynthetic en- 
zymes is evidence that a portion of the 
probe is introduced into the endo- 
plasmic reticulum and/or the Golgi ap- 
paratus, which are the principal sites of 
lipid biosynthesis in cells. Of all the 
NBD-labeled lipids tested, only NBD- 

NBD- Triglyceride 
NBD- Digiyceride 

- NBD- Fatty acid 



2 3 4 5 6 

Fig. 23. Metabolism of NBD-labeled phospholipids by V79 Chinese hamster lung fibroblasts. 
Lipids were extracted from cells after incubation with vesicles containing NBD-labeled phospho- 
lipids. Individual lipid classes were separated by thin-layer chromatography, and the fluorescent 
lipids were visualized by illumination with ultraviolet light. Lanes 1 and 6: Lipid standards (top to 
bottom) NBD-diglyceride, NBD-aminocaproic acid, NBD-PC, and NBD-PA. Lanes 2 and 4: Fluores- 
cent lipid products of cells incubated with NBD-PA for 1 h at 2°C (lane 2) and for 30 min at 37°C 
(lane 4). Lanes 3 and 5: NBD-PC incorporated into cells incubated for 1 h at 2°C (lane 3) and for 30 
min at 37 °C (lane 5). 



PA penetrated to the innermost mem- 
branes of the cell (such as the nuclear 
membrane). It is also the only lipid we 
found that when taken up by the cells 
was metabolized during the one-hour 
incubation period. 

Currently, we are elucidating the 
mechanisms by which NBD-PA is in- 
ternalized. It is tempting to suggest 

that the movement of NBD-PA is facil- 
itated by enzymes or carrier proteins 
within the cell. If so, the presence of 
such gene products may indicate an im- 
portant link between the movement of 
phosphatidic acid between intracellular 
membranes, and the regulation of lipid 
metabolism in cells. 

DURING Caenorhabditis elegans SPERMATOGENESIS 

S. Ward, D. Burke, T. Roberts, and E. Hogan 

Our laboratory is trying to learn how 
the morphology of a cell is specified 
genetically. We would like to know 
which gene products directly determine 
cell shape, how these gene products act, 
how they become localized to specific 
parts of the cell, and how they deter- 
mine the sequence of events in morpho- 
genesis. Our approach to this problem 
is a genetic one: We isolate a large 
number of mutants that affect the mor- 
phogenesis of a single cell type and 
then analyze these mutants anatom- 
ically and biochemically to learn the 
normal function of the mutated gene. 
The cell that we have chosen to study is 
the amoeboid sperm of the nematode 
Caenorhabditis elegans (Year Book 77, 
78, 79). 

This year, we have found that sper- 
miogenesis, the conversion of spherical 
sessile spermatids to bipolar motile 
spermatozoa, can be induced synchron- 
ously in large populations of purified 
sperm. This has allowed us to study in 
detail the metabolic and macromolecu- 
lar changes going on during this 
dramatic event in cell morphogenesis. 
These studies have shown that sper- 
miogenesis can be initiated either by an 
increase in intracellular pH or by pro- 
tease digestion of the cell surface. New 
protein synthesis is not necessary, and 
modifications of major sperm proteins 
by proteolysis, phosphorylation, or 

methylation have not been found. Thus 
the morphological changes occurring 
during spermiogenesis appear to be 
caused by the rearrangement of preex- 
isting macromolecules without their al- 

Both actin and the major sperm pro- 
tein—an abundant, small, soluble, ba- 
sic sperm protein— become confined to 
the pseudopod, while the cell organelles 
remain in the cell body. As described in 
Year Book 79, the plasma membrane 
also becomes polarized so that the 
pseudopodial membrane flows from tip 
to base, while the cell-body membrane 
is immobile. It is still not apparent 
what guides these intracellular and sur- 
face rearrangements. The cell lacks a 
conventional cytoskeleton of micro- 
filaments and microtubules, and so it 
must have a novel mechanism control- 
ling cellular asymmetry. 

The isolation of mutants altered in 
sperm development has continued. We 
have now accumulated enough mu- 
tants to estimate that there may be 
only 25-30 gene products necessary for 
spermiogenesis but not required by 
other nematode cells. Morphological 
defects were found in the sperm of six 
of nine mutants examined so far. The 
small number of sperm-specific genes 
and the large fraction of them involved 
in morphogenesis encourages the hope 
that continued biochemical analysis of 



the mutant defects will lead to an 
understanding of the molecular mech- 
anisms of sperm morphogenesis. 

An Increase in Intracellular pH 
Can Induce Sperm iogenesis 

S. Ward and E. Hogan 

In Year Book 78 and 79, we described 
the activation of spermiogenesis in 
vitro by the treatment of spermatids 
with the ion-transporting drug monen- 
sin. This ionophore can transport Na + 
and K + ions through membranes and 
exchange them for H + ions. Suspecting 
that H + transport was causing the ac- 
tivation by increasing the intracellular 
pH, we tested whether weak bases able 
to penetrate and raise the intracellular 
pH would also activate spermiogene- 
sis. We found that both ammonia and 
triethanolamine could mimic the effect 
of monensin and activate spermatids to 
spermatozoa (Fig. 24). The activation 
was dependent on the concentration of 
unprotonated amine (the form that 
would be expected to penetrate cells). 
Treatment of cells with triethano- 


-i 80 


r 'TV 


/ " ■* 

I- 60 

/ ' N 



/ / 

8 40 


55 20 


A . 


— 9f\ 

i i i 


20 30 40 


A] mM 

Fig. 24. Activation by triethanolamine (TEA). 
Sperm medium was adjusted to pH 7.8, and 
varying concentrations of TEA were added. The 
abscissa shows the calculated concentration of 
unprotonated TEA. The ordinate shows the % 
spermatozoa. The solid line is in normal sperm 
medium; the dotted line is in sperm medium with 
all added Na + and K + replaced by choline, ex- 
cept for 25 mM Na + added to adjust pH. The 
square shows the same medium without Na + . 

lamine caused a rapid increase in in- 
tracellular pH, from 7.1 to 7.6, pre- 
ceding the initiation of spermiogenesis 
(Fig. 25). The pH must remain elevated 
in the cell for five minutes for all cells to 
complete spermiogenesis. When sper- 
matids are treated with monensin, a 
similar increase in intracellular pH 
precedes spermiogenesis. 

These results show that the mecha- 
nism of action of both monensin and 
weak bases involves an increase in in- 
tracellular pH that somehow activates 
the process of spermiogenesis. Ele- 
vated pH levels cause or accompany 
developmental changes in many other 
cells. For example, an increase in pH in 
sea urchin eggs causes the stimulation 
of macromolecular synthesis after fer- 
tilization, and in protozoa, changes in 
intracellular pH accompany the initia- 
tion and termination of DNA replica- 

A change in pH is not the only way to 
initiate spermiogenesis, since protease 
treatment can induce spermiogenesis 
without altering intracellular pH (Fig. 

4 6 8 

TIME (min) 

Fig. 25. Time course of intracellular pH 
changes and activation in response to triethanol- 
amine (TEA). All experiments are in sperm me- 
dium pH 7.75; TEA was added just after taking 
T = point, (a) Unprotonated TEA at 30 mM. 
(b) Unprotonated TEA at 10 mM. Dashed line, 
pH; solid line, % control spermatozoa. 



26). All attempts to demonstrate pro- 
tease action accompanying weak-base- 
induced spermiogenesis have been 
negative, so it is not known how the 
protease and pH-initiated activation 
are related. Both mechanisms result in 
apparently identical spermatozoa of 
normal morphology and motility. 

Few Macromolecular Changes 
Accompany Spermiogenesis 

S. Ward and E. Hogan 

Since from 10 7 to 10 8 isolated sper- 
matids can be activated to spermato- 
zoa synchronously by triethanolamine 
treatment, we have begun to examine 
some of the biochemical changes 
accompanying this morphological dif- 
ferentiation. The only protein syn- 
thesis detected in spermatids or sper- 
matozoa is mitochondrial, the sperm 
have no ribosomes, and this mitochon- 
drial synthesis appears unnecessary for 
spermiogenesis. Energy metabolism by 
mitochondria is absolutely necessary 
for both spermiogenesis and sperm 
motility. At sub-micromolar concentra- 
tions, oligomycin, a drug that blocks 
oxidative phosphorylation, completely 
prevents spermiogenesis. Drugs that 
interfere with electron transport or un- 
couple electron transport from phos- 

phorylation also prevent spermiogene- 
sis when used at high concentrations. 

No RNA or DNA synthesis has been 
detected in sperm. Lipid biosynthesis 
is stimulated by spermiogenesis, but 
the pattern of lipids synthesized is 
unchanged from spermatids to sperma- 
tozoa. Similarly, protein phosphoryla- 
tion is stimulated during spermiogen- 
esis, but the pattern of proteins phos- 
phorylated and the relative amount of 
phosphorylation is the same in sper- 
matids and spermatozoa, at least at the 
resolution of one-dimensional gels. The 
same is true for the pattern of protein 
methylations. We have been unable to 
detect proteolytic enzymatic activity 
in spermatids or spermatozoa, and the 
pattern of the 400 or so proteins 
visualized on two-dimensional gels, ex- 
cept for a few variable spots, is iden- 
tical for both spermatids and sper- 
matozoa. These observations are all 
negative in the sense that no changes 
were observed, and we are pursuing 
them further with more sensitive 
methods. Nonetheless, they establish 
that no major changes in macromole- 
cules cause or accompany the mor- 
phological transition from spermatids 
to spermatozoa. This implies that the 
difference between these cells is in the 
arrangement, not the composition, of 
their macromolecules. 

<* 7.3 

I li&fy 

in 7.0 

^ 6.9 



4 6 8 

TIME (min) 

- 40 


Fig. 26. Time course of intracellular pH and 
activation with pronase. Pronase at 200 Mg/ml 
was added to cells just after time point was 
taken. Triangles show the pH change for two dif- 
ferent experiments. Circles show the activation 
of the same batches of sperm. 

The Localization of the Major 

Sperm Protein During 


S. Ward and M. Klass* 

It has been found by M. Klass and D. 
Hirsh and by ourselves that sperm con- 
tain a single protein that is 15% of their 
total protein. It is a small protein, 
15,500 daltons; it is basic, pi = 8.5; it is 
soluble and not nuclear; its native form 
is a dimer. Using an antiserum pre- 

* Department of Biology, University of Hous- 



pared against this major sperm protein 
and purified by affinity chromato- 
graphy, we have found by immunocyto- 
chemistry that it is localized uniformly 
in the cytoplasm of spermatids but is 
concentrated in the pseudopods of 
spermatozoa (Fig. 27). When spermato- 
cytes and immature spermatids are 
stained, the protein is localized in small 
patches whose positions correspond ex- 
actly to these found for the structures 
known from electron microscopy as 
fibrous bodies. The fibrous bodies con- 
sist of 4.3 nM fibers in paracrystalline 

arrays. Such fibers are not seen in nor- 
mal spermatids and spermatozoa. 

These observations suggest that the 
major sperm protein can exist in at 
least two forms, fibrous and amor- 
phous. The localization of this protein 
to the pseudopod suggests that it may 
participate in motility, but how it does 
so is unknown. Whatever its functional 
role, this protein is rearranged during 
morphogenesis so that antiserum di- 
rected against it is a useful probe to 
detect its altered arrangements in mu- 
tant sperm. 

Fig. 27. MSP localization during sperm development, (upper) Spermatocytes showing dots of 
fluorescence in the periphery, (middle) A spermatid budding off a residual body showing patches of 
fluorescence in the spermatid, (lower) Two spermatids showing uniform fluorescence and a spermato- 
zoon showing fluorescence confined to pseudopod. 



Genetic Estimation of the Number 
of Sperm-Specific fer Genes 

D. J. Burke and T. M. Roberts 

Sperm morphogenesis involves the 
interaction of many gene products. 
Some of these products will be used 
during the development of other cells in 
the nematode; others will be sperm- 
specific. By selecting mutants that 
have defective sperm but still grow, 
produce eggs, and behave normally, we 
should identify those genes whose 
products are sperm specific. How many 
such genes are there? 

Our mutant collection contains 35 
fertilization-defective (fer) mutants that 
have defective sperm but are otherwise 
normal. These mutants define at least 
16 different genes. We have used two in- 
dependent methods to estimate that 
these genes are representative of ap- 
proximately 30 total fer genes that are 
sperm specific. 

The first method is to determine the 
frequency of fer mutants that occur af- 
ter a standard amount of mutagenesis. 
This was done by raising individual Fl 
progeny of mutagenized hermaphro- 
dites and examining their F2 progeny 
for fer mutants. When young adults 
containing a full complement of sperm 
are mutagenized and a small number of 
Fl animals raised from each parent, 
these Fl animals are the product of two 
independently mutagenized gametes. 
In one experiment, 534 Fl animals 
were picked, and all their F2 progeny 
inspected for the fer mutant pheno- 
type. Of the 65 candidates identified, 
13 segregated a fer mutation in normal 
3:1 Mendelian ratios after two succes- 
sive backcrosses. The 534 Fl animals 
represent 1068 gametes. So the muta- 
tion rate per gamete is 13/1068 = 0.012. 
Several different laboratories have 
measured the average mutation rate 
per gene either by using a sex-linked 
lethal assay, correcting for double 
events, or by collecting many mutants 
in a single gene. Since identical condi- 

tions for mutagenesis were used in all 
experiments, the number of fer genes 
can be estimated by the ratio of the 
mutation rate for all fer genes (.012) to 
the average mutation rate per gene (5 X 
10 ~ 4 ). This method estimates that about 
24 genes can be mutated to produce a 
fer phenotype. (The calculation is based 
on the assumption that all mutants 
generated by the mutagenesis were 
identified, so 24 is probably an underes- 

The second method of estimating 
gene numbers is to determine the 
number of genes with multiple alleles. 
After backcrosses, mutants are assigned 
to linkage groups, tested for comple- 
mentation to existing mutants on the 
same chromosome, and mapped by 
three factor crosses. A summary of the 
genetic and phenotypic characteriza- 
tion of our mutant collection is shown 
in Table 1. The phenotypes of some of 
these mutants were described in Year 
Book 78 and 79. Many of our recent 
isolates identify alleles of previously 
defined genes. If the number of fer 
genes were very large, new mutants 
would have identified mostly new 
genes. From the actual distribution of 
the new alleles in previous genes, the 
total number of fer genes can be 
estimated. Assuming that all genes 
have an equal probability of being 
mutated, the frequency of multiple 
alleles should follow a Poisson distri- 
bution. It does so if fer-1, which ap- 
pears to be a hypermutable gene, is ig- 
nored. The best fit to the distribution is 
obtained if the total number of fer 
genes is 30. 

This second estimate is in close 
agreement with the first. Although 
both assume that the mutability of 
each gene is similar, their agreement 
lends confidence to our estimate that 
only 30 sperm-specific genes are nec- 
essary for sperm fertility. This is a 
small enough number that we should be 
able to identify all these genes by muta- 
tion. It indicates that the process of 


TABLE 1. Summary of Fertilization-Defective Mutants* 





fer-1 I 



fer-2 IV 

fer-3 1 




fer-5 III 


fer-6 I 

fer-7 I 

fer-10 IV 

fer-14 X 


fer-15 II 


let-9 X 












Sperm Phenotype 

Short pseudopods, no MO fusion, undirected membrane movement. 

Short or aberrant branched psuedopods. Perinuclear tubules. 

Some short pseudopods. Perinuclear tubules. 

Immotile pseudopods. Perinuclear tubules. 

Spermatids with inclusions. 

Spermatids with fibrous bodies. 

Normal motility, poor adhesion to spermatheca. 

Normal motility and morphology. No interaction with egg. 

Normal spermatid morphology. No response to mating or monensin. 

*A11 strains make sperm in near-normal numbers but lay unfertilized oocytes with few or no fertile 
eggs. The oocytes can be fertilized by wild-type males, so the sperm must be defective. Abs shows 
the number of absolute alleles, TS the number of temperature sensitive alleles. Unnamed genes have 
not been precisely mapped but they have been tested for complementation and do represent new 
genes. Eleven additional strains have been assigned to linkage groups but have not been tested yet 
for complementation. Six of these additional strains were recently sent to us from other laboratories. 

sperm morphogenesis cannot be hope- 
lessly complex and encourages the 
hope that genetic dissection of this pro- 

cess will make it possible to recon- 
struct the molecular mechanisms of 
cellular assembly. 


N. L. Nivera and P. J. Gear hart 

One goal in molecular immunology is 
to understand the mechanisms respon- 
sible for antibody gene diversification. 
Insights into antibody diversity have 
been obtained by comparing sequence 
information from immunoglobulin genes 
and proteins. Our current level of un- 
derstanding may be summarized as 

(1) Antibody molecules are composed 
of heavy (H) and light (L) chains, each 
of which contains variable (V) and con- 
stant (C) regions. (2) Comparative anal- 
yses of many variable regions have 

demonstrated that amino acid sequence 
variability is unevenly distributed 
throughout the V regions. The antigen- 
binding site is composed of extremely 
variable sections, termed hypervari- 
able regions (three for each chain), while 
nonbinding portions, termed the frame- 
work regions, are more highly con- 
served. The hypervariable regions fold 
to constitute the walls of the antigen- 
binding site. (3) The V regions fall into 
sets of closely related sequences called 
groups. (4) The V region of the light 
chain (V L ) is encoded by two distinct 



gene segments— V L (encoding amino 
acids 1-95), and J L (joining, encoding 
amino acids 96-108). The V region of 
the heavy chain (Vh) is encoded by 
three distinct gene segments— V H (en- 
coding amino acids 1-95), D (diversity, 
encoding amino acids 96-99), and J H 
(encoding amino acids 100-113). The J L 
gene segments encode only a small por- 
tion of the third hypervariable region of 
the light chain, whereas the D and J H 
gene segments encode most of the third 
hypervariable region of the heavy 

Four mechanisms are used to gener- 
ate V-region diversity. (1) There are at 
least a hundred germline V L and V H 
gene segments, indicating that germ- 
line diversity is extensive. (2) There are 
four functional J gene segments on the 
light and heavy chains, and it appears 
that any J gene segment may be joined 
with any V gene segment on the kappa 
light chain. Likewise, the combinatorial 
joining of V H , D, and J H gene segments 
is an important mechanism for amplify- 
ing diversity in the third hypervariable 
region of the heavy chain. (3) The site- 
specific recombination mechanism that 
joins the V L and Jl, and the V H , D, and 
J H gene segments may join particular 
gene segments at different points in 
their sequences, thus generating hy- 
brid codons as well as codon insertions 
or deletions. Thus, alternative joining 
points generate variation at these junc- 
tions during the act of gene-segment 
joining. (4) Somatic mutation generates 
diversity in the hypervariable and 
framework regions. 

To assess the patterns of variable- 
region diversity for a simple hapten, we 
have used the hybridoma technique to 
obtain homogeneous antibodies from 
BALB/c mice immunized against phos- 
phorylcholine. The work was performed 
in collaboration with Nelson Johnson, 
Richard Douglas, and Leroy Hood at 
the California Institute of Technology. 
We report here the amino acid sequence 
analyses of the N-terminal V H and V L 
regions from 27 hybridoma proteins as 

well as the complete sequence analysis 
of nine of these V H regions. There are 
more V H and V L regions participating 
in the phosphorylcholine response than 
can be encoded directly by germline Vh 
and V L gene segments. Moreover, the 

V regions from IgG antibodies are con- 
siderably more variable than those 
from their IgM counterparts. These ob- 
servations raise the possibility that a 
somatic mechanism for V-region diver- 
sification produces greater diversity in 
IgG than it does in IgM antibodies. 

Somatic Mutation Generates Diversity 

in the Heavy and Light Chains of 

A n ti-Phosphorylcholine A n tibodies 

To determine if somatic mutation oc- 
curs to increase diversity, it is neces- 
sary to know the number and sequences 
of the germline genes and to compare 
them to the protein sequences. The 
heavy chains of anti-phosphorylcholine 
antibodies come from one group of V H 
gene segments called the T15 V H group. 
Although there are four germline genes 
in this group, one gene is used almost 
exclusively to generate 98% of the 
heavy chains that have been sequenced 
(S. Crews et al, Cell, 25, 59-66, 1981). 
As seen in Fig. 28, the amino acid se- 
quences of 27 proteins are identical to 
the prototype sequence that is encoded 
in the germline, and there are 12 
variant sequences. (Hybridoma HPCG15 
is an exception that is derived from a 
rarely used second-germline gene.) The 
amino acid analysis of 19 complete Vh 
regions is shown in Fig. 29. Within the 

V segment encoding amino acids 1-95, 
ten antibodies are identical to the pro- 
totype germline sequence, and nine, an- 
tibodies have substitutions. Therefore 
most of the proteins are identical to the 
germline gene, and the variant proteins 
must arise by somatic mutation. 

Kappa light chains for anti-phospho- 
rylcholine antibodies come from three 
groups, known as the M167, T15, and 
M603 V L groups. There is only one 
germline V L gene segment encoding the 



£ £ + I + l l + I I + + +++++III 

«o o 

O N 

2 - S "• 




I* o _ H 

oo » 
8 o j- 


O O °- 

O O _ o 


£ S25 

T 1 










•X f>- lOO-N 

.i, .$ a> cu 

C w s ° 

>> o 
•° o'ScS 

■s 5 s o 

S.S = « 

CO ,Q <C, CD 

^ s ~ 

■B. 3 

S -O 

5b c -2 J3 

-Q SB _ 

a «> s 


•s '-3 

.i JO 













































V H Region 

10 20 X 







r i fT 

-Y— S 












































Fig. 28. V H regions of anti-phosphorylcholine 
antibodies. Hybridoma and myeloma proteins 
are listed on the left. Amino acids are noted by 
the one-letter code. HV1 designates the first hy- 
pervariable region. A solid line indicates identity 
with the prototype sequence. Spaces signify un- 
determined residues. 

M167 group (E. Seising and U. Storb, 
Cell, 25, 47-58, 1981). The data in Fig. 
30 show that nine proteins are identical 
to the germline-encoded sequence over 
the first 38 amino acids. Six proteins 
are different. Somatic mutation must 
have generated this diversity in Ml 67 
V L regions. It is not known how many 
germline genes code for light chains in 
the T15 and M603 V L groups. 
However, the sequences of proteins in 
these groups (Fig. 31) indicate that 
each group has a prototype sequence 
that is repeated identically many 
times. We will assume that proteins 
differing from the prototype sequence 
are variant proteins. 

The pattern of amino acid diversity 
occurring in the V H and V L regions 

shown in Figs. 28-31 is summarized as 
follows: (1) mutation occurs in both the 
framework and hypervariable regions, 
(2) most amino acid substitutions can 
be explained by single nucleotide 
changes, and (3) as many as eight sub- 
stitutions have occurred in one V H re- 
gion (Ml 67 in Fig. 29), indicating that 
somatic mutation can generate exten- 
sive diversity. 

IgG Antibodies to Phosphorylcholine 

Exhibit More Diversity Than Their 

IgM Counterparts 

None of the V H segments from 13 
IgM antibodies show variation from 
the prototype sequence (Figs. 28 and 
29) and only one of 14 V L regions -from 
IgM antibodies is a variant (Figs. 30 
and 31). In marked contrast, 13 of 18 
V H segments from IgG antibodies are 
variants, and 8 of 17 V L regions from 
IgG antibodies are variants. Thus, only 
one of 14 IgM antibodies has an amino 
acid substitution in either the heavy or 
light chains, whereas 14 of 17 IgG anti- 
bodies have amino acid substitutions in 
the heavy or light chains. Somatic mu- 
tation also occurs in IgA antibodies: 4 
of 9 antibodies have amino acid substi- 
tutions in either the heavy or light 

Cellular and recombinant DNA stud- 
ies have suggested that B cells initially 
produce IgM and then switch to pro- 
duce IgG or IgA after antigen stimula- 
tion. Thus, the greater diversity of the 
IgG hybridomas must arise either from 
infrequent variant V H gene segments 
in IgM-producing cells or from variant 
V H gene segments that are generated 
in IgG-producing cells. Two important 
questions arise. First, how and when 
are the variant sequences generated? 
Second, what selective pressures could 
give rise to the differences between 
IgM and IgG diversity? 

Somatic mutation may occur by one 
of two mechanisms, somatic recombi- 
nation or point mutation. Recent data 
on the nucleotide sequences of the 



Diversity in MI67 V|_ Regions 

10 20 30 















M5II K Ig A 

MI67 IgA 

Fig. 30. M167 V L group of anti-phosphoryl- 
choline antibodies. The prototype sequence is 
shown at the top. 

genes in the T15 V H group indicate 
that somatic recombination among 
germline genes does not generate the 
variant proteins in Fig. 29 (S. Crews et 
al, Cell, 25, 59-66, 1981). The data 
favor a point-mutation mechanism 
wherein nucleotide substitutions may 
be introduced by error-prone DNA rep- 
lication. Somatic mutation appears to 
occur after V-J joining, since IgM anti- 
bodies show no variation from germline 
gene sequences. 

V. Region 











































DG- IgA 

F IgA 

Fig. 31. T15 and M603 V L groups of anti- 
phosphorylcholine antibodies. The prototype se- 
quence for each group is shown. 

Two points are of special interest 
with regard to mechanisms of somatic 
mutation. First, activation of the so- 
matic diversification mechanism may 
occur in conjunction with the switching 
of heavy chain classes. As the somatic 
diversification mechanism generates 
variants in V L as well as Vh gene seg- 
ments, it must be activated in a trans 
fashion on different chromosomes. Sec- 
ond, somatic mutation may occur 
throughout the lifetime of individual B 
cells. Accordingly, the oldest B cells 
(those having undergone the most 
clonal expansion) accumulate the most 

Two types of biological selection may 
cause cells secreting variant IgGs to be 
selectively expanded over cells secret- 
ing germline-encoded IgM antibodies. 
(1) Antigen-driven selection. B-cell 
clones expressing variant antibodies 
may be expanded because they have a 
higher affinity for antigen than do B 
cells expressing prototype antibodies. 
As IgG-producing cell populations may 
have undergone more antigen-driven 
selection than IgM populations, IgG 
antibodies should show more variabil- 
ity than IgM antibodies. However, 
IgM antibodies have the same affinity 
for phosphorylcholine as IgG anti- 
bodies (data not shown). Therefore, 
among the antibodies we have studied, 
somatic mutation has not produced an- 
tibodies with higher affinity for anti- 
gen. (2) Idiotype selection. Selection of 
infrequent variants may occur by idio- 
type-specific regulation. For example, 
if idiotype-specific suppression controls 
the level of predominant clones express- 
ing the prototype sequences, B cells 
with altered sequences may be selec- 
tively expanded. However, recent stud- 
ies by S. Pierce and H. Kohler in col- 
laboration with us suggest that T cells 
cannot distinguish between the idio- 
types on germline vs. variant anti- 

In summary, the immune system 
seems capable of generating an enor- 
mous amount of variability by the so- 



matic mutation of a limited number of 
germline genes. An animal may first 
respond to antigen with IgM, anti- 
bodies using a finite number of germ- 
line V genes. After antigen stimulation, 
increased diversity may be introduced 
by mutation during prolonged clonal 
expansion of IgG- or IgA-producing 
cells, or perhaps during the class switch 
itself. The purpose of somatic mutation 
may be to introduce random substitu- 
tions throughout V regions in order to 
increase the repertoire of antibodies 
available to bind antigen. Only a small 
minority of the variants may have a 
higher affinity for antigen. We are cur- 
rently isolating several V-region genes 
of the M167 V L group from hybridoma 
cells that have undergone somatic mu- 
tation (see Fig. 30). The nucleotide se- 
quences of the V region and noncoding 
flanking regions will be compared to 
the germline sequence to determine 
how far mutation occurs in the DNA. 
This study may suggest a mechanism 
for somatic mutation. 

Most B Cells That Have Switched 

Surface Immunoglobulin Isotypes 

Generate Clones of Cells That Do Not 

Secrete IgM 

In addition to generating diversity in 
V genes, B lymphocytes undergo an 
unusual form of cell differentiation 
wherein they switch expression of 
heavy chain C regions. The order of 
genes encoding the constant regions of 
immunoglobulin heavy chains is 5'-C M , 
C 6 , C 7 3, C 7 i, C 7 2b» C 72a , C e , C a -3'. To try 
to determine whether switching of the 
expression of heavy chain genes is irre- 
versible and occurs in a direction from 
5' to 3' within a cell line, our lab, in col- 
laboration with John Cebra (University 
of Pennsylvania), has isolated B cells 
bearing particular membrane isotypes, 
stimulated them to generate clones, 
and assayed daughter cells for secreted 
antibody. B cells with and without sur- 
face IgM and IgD, and B cells with and 
without surface IgG, were isolated by 

fluorescence-activated cell sorting. 
Fractionated cells were individually 
stimulated with phosphorylcholine in a 
splenic-fragment assay, and the anti- 
body secreted by clonal progeny was 
assayed for several heavy chain iso- 
types by radioimmunoassay. The re- 
sults show that (1) B cells bearing IgM 
and IgD can produce individual clones 
of cells secreting three isotypes of anti- 
phosphorylcholine antibody— IgM, IgG, 
and IgA; (2) most B cells expressing 
surface IgG generate clones which se- 
crete IgG but not IgM; and (3) most B 
cells from gut-associated lymphoid tis- 
sue that do not express surface IgM or 
IgD generate clones of cells which se- 
crete IgA but not IgM. Therefore, IgM- 
and IgD-bearing cells can give rise to 
progeny that secrete IgM, IgG, and 




unstained, sM.O 


stained, sG sG 

unsorted pot 


unsorted pos neg 




Fig. 32. Comparison of surface isotypes on B 
cells to IgM secreted by clonal progeny. Cells 
were separated according to membrane isotype 
by fluorescence-activated cell sorting, and stim- 
ulated with phosphorylcholine. The antibody se- 
creted by individual clones was then assayed for 
the IgM isotype. (sM, D pos) cells, with surface 
IgM and IgD; (sM, D neg) cells, without surface 
IgM and IgD; (sG pos) cells, with surface IgG; 
(sG neg) cells, without surface IgG. Ninety-five 
percent of the cells in the surface slgG-negative 
fraction had surface IgM. 



IgA. Once B cells switch to a different 
surface isotype, they appear restricted 
in isotype potential since they fre- 
quently generate clones of cells that do 
not secrete IgM (Fig. 32). These find- 

ings suggest that isotype switching is 
generally irreversible, and that iso- 
types are expressed in a 5' to 3' direc- 
tion that is consistent with the germ- 
line gene order. 


D. Chaleff, J. Mauvais, S. McCormick, M. Shure, S. Wessler, and N. Fedoroff 

Transposable controlling elements in 
maize have been subjected to extensive 
genetic analysis, chiefly by B. McClin- 
tock, M. Rhoades, and R. A. Brink. Mc- 
Clintock's work on controlling elements 
in maize is summarized in successive 
Year Books beginning with Year Book 
45 (1946). Controlling elements have 
been identified at many loci in different 
biosynthetic pathways in maize, includ- 
ing those for starch, chlorophyll, and 
anthocyanin pigments. As is true for 
other transposable elements, insertion 
of maize controlling elements frequently 
inactivates or severely restricts expres- 
sion of a locus. Insertion mutants gen- 
erally display a recessive mutant pheno- 
type. Such insertions are quite unstable 
in maize, giving rise to both a pattern 
of somatic reversion and a high fre- 
quency of genetically stable mutants, 
including revertants. It has been ob- 
served that a controlling element can 
transpose away from a mutant locus. 
In some cases, the mutant locus does 
not revert to the normal phenotype, but 
continues to give the pattern of somatic 
and germline back mutation character- 
istic of the original mutant. However, 
somatic reversion occurs only in the 
presence of the controlling element. In 
its absence, the allele behaves as a sta- 
ble recessive mutant. The transposable 
element therefore appears to produce a 
trans-acting substance necessary for 
somatic and germline reversion. This 
trans-acting function was initially des- 
ignated the mutator. It has subse- 
quently been shown to be required for 
transposition of both the controlling 

element and the so-called receptor, 
which is the name used to designate the 
insertion sequence at the locus. The 
mutator should therefore perhaps be 
termed the transposition function. The 
receptor sequence at the locus is cis- 
acting, affecting only the expression of 
the locus to which it is closely linked. 
Thus a controlling element system in 
maize generally comprises a trans-act- 
ing controlling element and the locus 
altered by insertion of a cis-acting re- 
ceptor sequence. Because such a two- 
element system has been observed to 
arise after transposition of a control- 
ling element away from a locus, it has 
been conjectured that the receptor com- 
prises a defective transposable ele- 
ment. Since many of the transposable 
element mutants in maize were two-ele- 
ment systems at the time of their dis- 
covery, the close relationship between 
the receptor and the transposable ele- 
ment will remain conjectural until these 
sequences are isolated and subjected to 
molecular analysis. 

Maize transposable element mutants 
alter, in a very specific way, a gene's 
developmental pattern of expression. 
Different insertion mutants of the 
same locus show distinctly different 
patterns of response to the controlling 
element. These differences are defined 
by the frequency and time of occurrence 
in development of somatic back muta- 
tion and the intensity of gene expres- 
sion in revertant sectors. There also ex- 
ist mutants of the transposable element 
which alter these parameters. In addi- 
tion, there are controlling element mu- 



tants which are expressed in tissue- 
specific ways, suggesting that the 
activity of the transposable element 
can be regulated by compounds differ- 
entially distributed between tissues or 
within a tissue. 

Several different transposable ele- 
ment systems have been identified in 
maize. Of these, the most extensively 
studied are the Activator-Dissociation 
(Ac-Ds) and Suppressor-mutator (Spm) 
controlling element systems. To gain 
access to these systems at the molecu- 
lar level, we have concentrated our ini- 
tial efforts on several structural loci for 
which the gene products are known, 
without regard to which system exerts 
control. These include two loci in the 
anthocyanin pigment biosynthetic 
pathway and two loci encoding genes 
affecting starch biosynthesis. Those in- 
volved in anthocyanin pigment synthe- 
sis are the Bronze (Bz) locus, which en- 
codes a flavonoid glucosyl transferase, 
and the C2 locus, which encodes the 
flavanone synthetase. The starch bio- 
synthetic enzymes that we are focusing 
on are sucrose synthethase, which is 
the product of the Shrunken (Sh) locus 
on chromosome 9, and a starch particle- 
bound glucosyl transferase encoded by 
the Waxy (Wx) locus on the same chro- 
mosome. Each of these loci is impor- 
tant in the context of this study, and 
each holds both advantages and disad- 
vantages. The loci in the anthocyanin 
biosynthetic pathway are important 
because mutations affecting both the 
controlling element and the responsive 
locus are readily identified visually. 
This is so because the anthocyanin pig- 
ments are expressed in the surface 
layer (aleurone) of the kernel. Many 
thousands of kernels can be scanned 
readily for mutants. Because the em- 
bryo and endosperm, of which the aleu- 
rone is the outermost layer, are geneti- 
cally identical, phenotypically mutant 
kernels generally give mutant plants, 
facilitating propagation of mutants. 
Unfortunately, the enzymes involved 
in anthocyanin biosynthesis are not 

abundant. Both the Bz and C2 gene 
products appear to comprise substan- 
tially less than 0.1% of the total kernel 
protein, which complicates their purifi- 
cation and makes cloning the corre- 
sponding mRNAs formidable. Muta- 
tions affecting the quantity or quality 
of endosperm starch are less readily 
visualized. However, these enzymes ap- 
pear to be quite abundant, which makes 
cloning the encoding mRNAs (and 
genes) quite feasible. Progress made in 
the past year on these loci is summa- 
rized below. 

The Bronze Locus 

N. Fedoroff and J. Mauvais 

Purification of the glucosyl transfer- 
ase was described in Year Book 79. We 
have subsequently confirmed the iden- 
tity of the purified polypeptide asso- 
ciated with the Bz locus by analyzing 
proteins extracted from aleurone tissue 
of normal strains and of a strain be- 
lieved to be deleted for the Bz locus. We 
find that the polypeptide we have puri- 
fied on the basis of its ability to glyco- 
sylate flavonoids is indeed the polypep- 
tide missing from the protein pattern 
observed on a two-dmensional gel of 
aleurone proteins from the mutant 

Our present efforts are aimed at 
identifying the mRNA for the glucosyl 
transferase. Messenger RNA prepara- 
tions have been made from normal and 
mutant strains and the in vitro transla- 
tion products are being analyzed on 
two-dimensional gels and by immune 
precipitation using antibody to puri- 
fied glucosyl transferase. The major 
difficulty we have encountered is that 
the mRNA coding for the glucosyl 
transferase appears to be much less 
abundant relative to other mRNAs 
than is the protein relative to other al- 
eurone proteins in mature endosperm. 
This suggests that the protein is an ex- 
tremely stable one that accumulates 
over a long period of time. Since these 



experiments indicate that cloning the 
mRNA for the flavonoid glucosyl trans- 
ferase directly would be fairly difficult, 
we have expended some effort on the 
characterization of a McClintock- 
derived mutant, designated bz m4 , which 
behaves genetically like a deletion for 
the Sh locus and brings the element Ds 
into close juxtaposition with the Bz 
locus, rendering it unstable, as de- 
scribed earlier. In this mutant, recom- 
bination between what remains of the 
Sh locus and the Bz locus is reduced 
substantially, from about 2% to roughly 
0.06%. Further biochemical characteri- 
zation of this mutant had the objective 
of characterizing the Sh locus in this 
strain and appears below in conjunc- 
tion with other studies on the Sh locus. 

The C2 Locus 

S. McCormick 

On the basis of precursor feeding 
experiments, the C2 gene product is 
believed to act at an early step in the 
anthocyanin biosynthetic pathway 
(McCormick, Biochem. Genet, 16, 777- 
785, 1978). Hugo Dooner has recently 
shown (Genetics, 94, s29, 1980) that c2 
dosage in the aleurone of maize corre- 
lates with flavanone synthetase activ- 
ity, and that c2 aleurone has no flava- 
none synthetase activity. This suggests 
that c2 is the structural gene for flava- 
none synthetase. The c2 locus is of in- 
terest because it is the only gene in 
either the anthocyanin or starch bio- 
synthetic pathway with a known enzy- 
matic activity at which there is a muta- 
ble allele (c2 ml ) that has the entire Spm 
controlling element (regulator and re- 
ceptor) at the locus. 

Flavanone synthetase catalyzes the 
reaction between malonyl CoA and 
p-coumaryl CoA to form flavanone. 
This enzyme has been purified to homo- 
geneity from parsley cell suspension 
cultures (Hahlbrock et al., Eur. J. 
Biochem., 99, 89-96, 1979). A purifica- 
tion scheme for flavanone synthetase 

from immature maize aleurone tissue 
has been adapted from the purification 
procedure developed for the enzyme 
from parsley. The maize enzyme resem- 
bles the parsley enzymes in its chromat- 
ographic properties and instability. 
The first purification step is a 65-80% 
ammonium sulfate precipitation, which 
eliminates 94% of the protein. The am- 
monium sulfate precipitate is desalted 
in a Sephadex G-25 column and applied 
to a hydroxylapatite column. The bulk 
of the protein elutes at a phosphate 
concentration of less than 0.1 M, while 
flavanone synthetase activity elutes at 
0.15 M Na phosphate. The enzyme is 
further purified by chromatofocusing. 
The enzymatic activity elutes at a pi of 
5.6. Four polypeptides, ranging in size 
from 45 kD to 65 kD, were found to 
co-elute with flavanone synthetase ac- 
tivity. To identify which of these poly- 
peptides is associated with flavanone 
synthetase, proteins from C2 and c2 
aleurone tissue are being compared by 
two-dimensional polyacrylamide gel 

The Waxy Locus 

M. Shure and S. Wessler 

Of the many loci found in association 
with the controlling element systems 
of maize, the Waxy locus has many ad- 
vantages which make it attractive for 
biochemical and molecular analysis. It 
had been reported by others that this 
locus encodes a starch-granule-bound 
UDP-glucose- starch glucosyl transfer- 
ase (Nelson et al, Biochem. Biophys. 
Res. Comm., 9, 297, 1962; Akatsuka et 
al, J. Biol. Chem., 241, 2280, 1966). The 
starch in both endosperm and pollen 
grains comprises only amylopectin in 
u;;c -recessive mutants, while the starch 
of the corresponding wild-type tissue is 
a mixture of amylose and amylopectin. 
The dominant and the recessive mu- 
tant phenotypes can be conveniently 
distinguished by staining the endo- 
sperm or pollen grains with KI-I 2 rea- 



gent. Another attractive feature of this 
locus is that it is well characterized 
genetically. A fine-structure map of the 
Waxy locus has been constructed by 
Nelson (Nelson, Genetics, 60, 507, 
1968). Because the Wx phenotype can 
be determined by staining pollen 
grains, Nelson was able to measure 
intra-allelic recombination frequencies. 
In this manner, he was able to map 
numerous recessive {wx) alleles in addi- 
tion to several of the known mutable 
alleles (wx m ) of this locus. These stud- 
ies demonstrate that several of the con- 
trolling element mutations recombine 
on both sides with standard recessive 
mutations, indicating that the control- 
ling element mutations are within the 
Wx locus. Several controlling element 
mutants are available for analysis, in- 
cluding some with Ac at the locus. 

Investigations of the Wx locus have 
therefore been undertaken with the 
goal of studying the mutable alleles at 
the molecular level. In initial studies, it 
was found that starch granules pre- 
pared from immature endosperms, iso- 
lated from seeds harvested 27 days 
after pollination (27 DAP), contained a 
limited number of proteins. One of the 
major polypeptides found in Wx 
granules has a molecular weight of 
55,000 daltons (55 kD) (Fig. 33b) and is 
absent from granules prepared from 
several wx strains (Fig. 33a). Subse- 
quently, it was found that the relative 
abundance of this protein was greater 
in granules prepared by differential 
centrifugation from younger endo- 
sperms (20 DAP) (Fig. 33c). In such a 
granule preparation, all of the UDP- 
glucose-dependent starch-synthesizing 
activity co-purifies with the 55-kD 
band, which in turn co-purifies with the 
granules. In the final granule pellet the 
approximate ratio of protein to dry 
mass, which is almost entirely starch, 
is 5 [iglmg. The difference between 
wild-type and mutant tissue is even 
more striking when the proteins ex- 
tracted from granules prepared from 
18-20 DAP endosperms are examined 

a b 

Fig. 33. SDS-polyacrylamide gel electropho 
retic analysis of starch-granule proteins. Pro- 
teins extracted from starch granules by boiling 
in 2% SDS are displayed in lanes (a), (b), and (c). 
The granules were from a wx strain (a), a Wx 
strain (b) (both harvested 27 DAP), and a Wx 
strain (c) (harvested 20 DAP). Lane (d) shows the 
55-kD protein after DEAE cellulose chromatog- 
raphy. The arrow indicates the position of the 
55-kD protein. 

in parallel on 2-D gels. The overall pat- 
terns are indistinguishable with the ex- 
ception of the 55-kD protein, which is 
present only in the wild-type sample. 

To establish the relationship between 
the enzymatic starch-synthesizing activ- 
ity and the 55-kD band, granules were 
prepared from triploid endosperms con- 



taining zero, one, two, or three doses of 
the wild-type Wx locus. It was found 
that the amounts of both the UDP-glu- 
cose-dependent starch-glucosyl trans- 
ferase activity and the 55-kD protein 
vary in proportion to the number of Wx 
doses. These results strongly suggest 
that the enzymatic activity and the 
55-kD band are in fact the same poly- 
peptide. In order to strengthen this 
argument, numerous attempts were 
made to dissociate the 55-kD protein 
and the enzymatic activity from the 
granules. Only harsh treatments suc- 
ceeded in dissociating the protein from 
the starch; unfortunately, these proce- 
dures resulted in a loss of enzymatic ac- 

In order to facilitate further biochem- 
ical studies of the Wx locus at the tran- 
scriptional and translational levels, it 
was decided to purify the 55-kD poly- 
peptide for use as an antigen in raising 
antiserum in rabbits. Because of the 
difficulty in dissociating an active poly- 
peptide, isolation of an inactive protein 
was undertaken. To this end, purified 
starch granules were solubilized in 8 M 
urea at 37°, and the aqueous phase con- 
taining all the granule proteins was col- 
lected following centrifugation, diluted 
to 4 M urea, and purified by DEAE cel- 
lulose chromatography at pH 8.0. The 
55-kD polypeptide eluted at approx- 
imately 30 mM NaCl in the presence of 
4 M urea. The gel profile of the peak 
protein fraction is shown in Fig. 33d. 
The 55-kD protein was further purified 
on an SDS-polyacrylamide gel. The 
amount of the 55-kD protein recovered 
is about 0.5% of the endosperm pro- 
tein. The gel-purified material was used 
to immunize rabbits. 

To more rigorously identify the 
55-kD protein as the product of the 
Waxy locus, the granule proteins of 
wx m strains were examined. Control- 
ling elements mediate somatic rever- 
sion of mutable alleles. In the absence 
of the controlling element, mutable 
alleles of Waxy are indistinguishable 
phenotypically from standard (wx) re- 

cessives. The presence of the control- 
ling element can be determined by 
staining individual endosperms with 
KI-I 2 reagent to detect the presence of 
revertant sectors. Correlated with the 
presence of the controlling element and 
the appearance of revertant sectors is 
the appearance of a protein of about 55 
kD. This protein is absent when the 
controlling element is absent (Fig. 34A, 
compare lane 4 with lanes 5 and 6; com- 
pare lane 8 with lanes 9 and 10). For 
both wx m7 , an allele responding to Ac 
(data not shown), and wx" 18 , the allele 
responding to Spm, the mobility of the 
new band was identical to that of the 
wild-type 55-kD protein (Fig. 34A, 
compare lanes 3 and 4). For wx m6 , an- 
other mutable allele responding to Ac, 
the new band present in samples de- 
rived from sectored endosperms has a 
slightly lower mobility than the wild- 
type band, suggesting that the control- 
ling element results in a mutation of 
the coding sequence (Fig. 34A, com- 
pare lanes 7 and 8). 

To determine whether the 55-kD pro- 
tein appearing in wx" 18 (+Spm) strains 
and the approximately 57-kD protein 
appearing in wx m6 (+Ac) strains were 
antigenic ally related to the wild- type 
protein, the granule proteins were elec- 
trophoretically transferred from SDS- 
polyacrylamide gels to diazobenzyloxy- 
methyl (DBM) paper and reacted with 
antiserum raised against the wild-type 
55-kD protein. The second antigen- 
combining site of the antibody was 
then labeled with 125 I-labeled 55-kD 
protein to locate the cross-reacting 
material. From the results shown in 
Fig. 34B, it is clear that only a single 
diffuse band (co-migrating with the 
55-kD protein) appears in channels 
wherein material from Wx, wx m8 
(+Spm), and wx m6 (+Ac) granule pro- 
teins was applied. No cross-reacting 
material was detected among the pro- 
teins extracted from granules of wx, 
wx m8 (—Spm), or wx m6 (—Ac) endo- 
sperms. These data support the previ- 
ous observation that the presence of 



123456 789 10 



wx m6 wx m8 


E E 

O O Q. CL 

* x < < * CO CO 


Fig. 34. (A) SDS-polyacrylamide gel electropfyoretic analysis of granule proteins from wx m 
strains. The major band in lane (1) is the 55-kD protein. Other lanes show the granule proteins from 
single kernels isolated 27 DAP. The strains represented are (2) homozygous wx m9 , no^4c; (3) Wx; (4) 
homozygous wx" 18 , Spm segregating in the cross, endosperms sectored; (5) homozygous wx m8 , Spm 
segregating in the cross, nonsectored endosperms; (6) homozygous wx m8 , no Spm in the cross; (7) 
wx mS /wx m6 , with Spm, no Ac in the cross, endosperms sectored; (8) wx m6 , Ac segregating in the 
cross, sectored endosperm; (9) wx m6 , Ac segregating in the cross, nonsectored endosperm; (10) wx m6 , 
no Ac in the cross. (B) Autoradiogram of a filter containing the protein samples described in the text 
and in (A), above, transferred to diazotized paper and probed with antiserum to the 55-kD protein 
and I 125 -labeled 55-kD protein, as described in the text. The labeled band co-migrates with a 55-kD 
protein standard. 

the new band correlates with the pres- 
ence of the controlling element that 
leads to somatic reversion. Further- 
more, these data show that the wx" 18 
protein and the larger wx m6 polypep- 
tide are both antigenically related to 
the wild-type 55-kD protein. The fact 
that the wx m6 allele (in the presence of 
Ac) results in the production of an 
altered polypeptide, which is antigeni- 
cally related to the wild-type protein, 
provides evidence that the Wx locus is 
indeed the structural locus for the 
55-kD band. 

Antiserum raised against the puri- 
fied 55-kD protein has been used to 
identify antigenically related material 
found among in vitro translation prod- 
ucts. When total endosperm mRNA de- 

rived from kernels harvested 20 DAP 
was translated, no major protein band 
of 55 kD was observed. However, a 
polypeptide of 60 kD was a major trans- 
lation product of wild-type mRNA, but 
was absent from two wx recessive prep- 
arations (Fig. 35, compare lanes 1, 2, 
and 3). This protein is substantially 
larger than the protein that had been 
identified tentatively as the Wx gene 
product. However, upon immune pre- 
cipitation of the in vitro translation 
products with antiserum to the 55-kD 
protein, the larger polypeptide was 
uniquely precipitated (Fig. 35, compare 
lanes 4 and 5), suggesting that it is a 
precursor synthesized but not pro- 
cessed in the in vitro rabbit reticulo- 
cyte translation system (Fig. 35, see 



3 4 5 6 7 8 9 

- 60k 

- 55K 

> * 



Fig. 35. SDS-polyacrylamide gel electrophoretic analysis of 35 S methionine-labeled proteins syn- 
thesized in a rabbit reticulocyte translation system. The proteins obtained using poly A + mRNA 
from (1) shbzwx, (2) ShBzwx, and (3) ShBzWx maize endosperm are shown. Lanes (4-9) represent 
immune-precipitable translation products using (4) ShBzWx mRNA and preimmune serum, (5) 
ShBzWx mRNA and antiserum to the 55-kD protein, (6) shbzwx mRNA and antiserum to the 55-kD 
protein, (7) ShBzwx mRNA and antiserum to the 55-kD protein, (8) ShBzWx mRNA and antiserum 
to the 55-kD protein prebound for 30 min with 7 /xg of unlabeled 55-kD protein, (9) ShBzWx mRNA 
and antiserum to the 55-kD protein with 125 I-labeled 55-kD protein added to the gel sample. 

lane 9 for the relative mobility of these 
two proteins). This was further sup- 
ported by the observation that no pro- 
tein was precipitated specifically by 
the antiserum from the translation 
products obtained from those mRNAs 
isolated from the vox recessive mutants 
(Fig. 35, compare lanes 6 and 7). In ad- 
dition, precipitation of the labeled 
60-kD polypeptide could be prevented 
by preincubating the antiserum with 
the unlabeled purified 55-kD protein 
(Fig. 35, lane 8). Since the mRNA of the 
60-kD protein represents one of the 

most abundant endosperm mRNAs, 
our current efforts are focused on clon- 
ing cDNA to this mRNA. 

The Shrunken Locus 

D. Chaleff, N. Fedoroff, and J. Mauvais 

The Shrunken locus is on the short 
arm of chromosome 9 and encodes su- 
crose synthetase. Sucrose synthetase 
is one of the most abundant soluble en- 
dosperm proteins and the controlling- 
element-associated alleles of this gene 



are likely to be among the first to be 
cloned. McClintock isolated three mu- 
table alleles of the Sh locus, designated 
sh m6233 , sh" 15933 , and sh m6258 . The origin 
of these mutations is described in Year 
Book 51 and 52. The sh m alleles were 
derived from a strain in which Ds was 
located between the C and Sh loci on 
the short arm of chromosome 9, but 
had no phenotypic effect on either. Ds 
appeared to be somewhat transposition 
defective in this strain, but gave rise to 
unidirectional deletions or rear- 
rangements having Ds as one end 
point. This strain gave rise to the sh m 
alleles as well as the bz m4 allele, which 
appears to be deleted for all or part of 
the Sh locus. The sh m alleles appear to 
have Ds (but not Ac) close enough to 
the Sh locus to produce the unstable re- 
cessive phenotype characteristic of 
controlling element mutations. It 
should be kept in mind, however, that 
the sh m alleles may have arisen either 
by transposition of Ds or by a deletion 
which brings Ds close enough to the Sh 
locus to affect its expression. 

The initial phases of cloning cDNA 
to sucrose synthetase mRNA in the 
plasmid pBR322 were reported in Year 
Book 79. The task proved unexpectedly 
difficult. Although long cDNAs were 
obtained from mRNA fractions enriched 
for sucrose synthetase mRNA, no su- 
crose synthetase cDNA clones were ob- 
tained when only the longest double- 
stranded cDNAs were used, despite a 
high yield of plasmids with long in- 
serts. Sucrose synthetase cDNA clones 
were obtained only when total cDNA 
was cloned. Approximately 20 sucrose 
synthetase cDNA clones were ulti- 
mately recovered. Of these, almost all 
were quite short, approximately 250- 
350 nucleotides. This represents only 
about 10% of the mRNA length, which 
is roughly 3000 nucleotides. Moreover, 
one of the clones proved very unstable, 
suggesting that some structural fea- 
ture of the mRNA precluded stability 
of the cDNA clones. The longest in- 
sert is 750 nucleotides and is the clone 

that we used in subsequent studies. 
The cDNA clone contains internal Pst 
I and Bgl II sites roughly 500 and 250 
nucleotides from one end. There are no 
Eco RI, Bam HI, Sst I, or Hind III 
sites within the cloned cDNA. 

Studies have been done on both the 
structure and the expression of the 
sucrose synthetase genes in mature en- 
dosperm from strains carrying the sh m 
alleles and the bz m4 allele. Studies on 
the expression of the Sh locus are some- 
what complicated by the existence of 
an enzyme which is extremely similar 
to the SA-encoded sucrose synthetase, 
but which is coded for by a distinct 
gene. The possibility that there exists 
a closely related sucrose synthetase 
gene was raised by the studies of 
Chourey and Nelson (Biochem. Genet., 
4, 1041, 1976). Subsequently, Chourey 
has isolated this enzyme from the bz m4 
strain (personal communication). That 
all or part of the Sh locus is deleted in 
the bz m4 strain is evidenced by the 
hybridization experiment represented 
in Fig. 36. DNAs from Sh and bz m4 
kernels were digested with the restric- 
tion enzymes Hind III and Bgl II, frac- 
tionated on an agarose gel, transferred 
to nitrocellulose filter paper, and hybri- 
dized with the cloned sucrose synthe- 
tase cDNA described above, labeled 
with P 32 by nick-translation. It is evi- 
dent that the two major bands hybrid- 
izing to cDNA in Sh DNA are absent in 
oz m4 dna. This is unequivocal evi- 
dence that at least part of the coding 
sequence is missing from the Sh locus 
in bz m4 . Minor bands are still present 
in both digests. There is a minor band 
at 1.1 kb in the Bgl II digest and one at 
5.1 kb in the Hind III digest. Since 
these bands are also observed in an 
shbz deletion mutant derived by J. 
Mottinger (Genetics, 64, 259, 1970), it 
appears likely that they represent cross- 
hybridization of the Sh cDNA probe 
with the related sucrose synthetase 
gene. The alternative explanation, that 
these bands represent a short, unde- 
leted fragment of the Sh coding se- 



Sh|bz m4 

Hind HI 
Sh|bz m4 






2.0- ** 



Fig. 36. Hind III and Bgl II restriction en- 
zyme fragments homologous to Sh cDNA in 
DNA from Sh and bz m4 strains. DNA from a 
Sh/bz m4 heterozygote and a bz m4 homozygote 
were digested with either Hind III or Bgl II, 
fractionated on an agarose gel, transferred to ni- 
trocellulose paper, and hybridized with cloned 
cDNA complementary to the S/i-encoded su- 
crose synthetase mRNA and labeled with 32 P by 

quence represented in the DNA clone, 
seems unlikely, since it would require 
that the two different deletion mutants 
are exactly coextensive. Chourey's 
characterization of the sucrose synthe- 
tase in the bz m4 deletion strain in- 
dicates that it is extremely similar to 
the SA-encoded enzyme in substrate 
specificity and specific activity, identi- 
cal to it in size (89 kD), and anti- 
genically indistinguishable (personal 
communication). We find that there is 
1-5% as much of an 89-kD polypeptide 
precipitable with sucrose synthetase 

antiserum in endosperm from the two 
sh deletion mutants examined as there 
is in Sh endosperm. The sucrose syn- 
thetase-like 89-kD polypeptide has 
been subjected to partial proteolytic di- 
gestion by S. McCormick. Comparison 
of the partial proteolytic products de- 
rived from this protein and the Sh- 
encoded sucrose synthetase showed 
that they are clearly distinct proteins, 
although probably related. Although 
indistinguishable on SDS-polyacryla- 
mide gels, they can be distinguished on 
native gels (Chourey and Nelson, Bio- 
chem. Genet, 4, 1041, 1976). Initial 
studies on the Sh-encoded sucrose syn- 
thetase in the sh m strains were carried 
out under denaturing conditions and 
were therefore limited in sensitivity by 
the co-immune precipitation and elec- 
trophoretic co-migration of the two su- 
crose synthetase enzymes. Nonethe- 
less, these experiments indicated that 
the amount of the SA-encoded sucrose 
synthetase present in endosperm from 
the sh m strains having Ds at the Sh 
locus but lacking Ac was no more than 
1-3% of the amount present in Sh en- 
dosperm. Experiments are currently 
being carried out under non-denaturing 
conditions to assess whether any Sh- 
encoded sucrose synthetase is present 
in these strains. 

In order to determine whether mRNA 
for the Sh-encoded sucrose synthetase 
is present in the sh m alleles having Ds 
at the locus, it was necessary to assess 
the relationship between the mRNAs 
encoding the two sucrose synthetases. 
That they are related is indicated by 
the results of the experiment displayed 
in Fig. 37. Poly A+ mRNAs from Sh 
and bz m4 endosperm tissue were 
hybridized to the Sh cDNA plasmid 
bound to a nitrocellulose filter. The 
hybridized mRNA was eluted and trans- 
lated in an in vitro rabbit reticulocyte 
protein-synthesizing system in the pres- 
ence of [ 35 S] methionine. The proteins 
were analyzed on an SDS-polyacryla- 
mide gel. Among the translation prod- 
ucts of the bz m4 mRNA hybridized to 





1 ^g 

Fig. 37. Proteins translated in vitro from 
mRNAs hybridizing to the Sh cDNA-contain- 
ing plasmid. The Sh cDNA plasmid was de- 
natured and bound to nitrocellulose. The filter- 
bound DNA was hybridized to poly A + mRNA 
from bz m4 or Sh endosperm tissue. The filters 
were washed and heated to 100° in H 2 to disso- 
ciate hybridized RNA. The RNA was collected 
by ethanol precipitation and translated in vitro 
in a rabbit reticulocyte cell-free system in the 
presence of 35 S methionine. The proteins were 
denatured and analyzed on an SDS-polyacryl- 
amide gel. The gel was then subjected to flu orog- 
raphy. The amounts of poly A+ mRNA and the 
strain from which it was derived are indicated. 
The arrow shows the position at which sucrose 
synthetase migrates. 

the plasmid was an 89-kD polypeptide 
(Fig. 37). Another indication that 
mRNAs are related but not identical 
was obtained when poly A + mRNAs 
from various Sh and mutant strains 
were fractionated on an agarose gel, 
transferred to nitrocellulose filter 
paper, and hybridized with Sh cDNA 
labeled with P 32 . The results of such an 
experiment are shown in Fig. 38. It can 
be seen that after prolonged autoradi- 

ography, a small amount of polynucleo- 
tide homologous to the Sh cDNA but 
slightly longer than the Sh-encoded 
mRNA is detectable in bz m4 . That the 
band is so faint is possibly due to the 
fact that these sequences cross-hybrid- 
ize rather poorly. The poor homology 
between the cross-hybridizing RNAs is 
evident from the observation that 
there is no mRNA in endosperm from a 
bz m4 strain able to form a hybrid with a 
portion of the Sh cDNA that is resis- 
tant to digestion with the single-strand 
specific nuclease SI. The results of 
such an experiment are shown in Fig. 
39. Various amounts of poly A + 
mRNA from Sh and bz m4 endosperms 
were hybridized to the 500-nucleotide 
Pst 1 fragment of Sh cDNA labeled 
with P 32 by nick translation. The hy- 
brids were digested with nuclease SI 
and denatured, and the nuclease-resis- 
tant fragment was analyzed on a poly- 
acrylamide gel. The fragment pro- 
tected by S7*-strain mRNA is slightly 
shorter than the intact cDNA frag- 
ment because the cloned cDNA has a 
homopolymer tail added during clon- 
ing. The protected fragment is indi- 
cated by the arrow in Fig. 39. The sen- 
sitivity of the assay is such that a 
homologous sequence present at a 2500- 
fold lower concentration than the 
amount of SA-encoded sucrose syn- 
thetase mRNA in an Sh strain could be 
detected. Under these conditions, 
neither a full-length nor a partially pro- 
tected fragment was observed when 
the hybridizing mRNA was derived 
from bz m4 (Fig. 39). Thus, the mRNAs 
encoding the two related 89-kD sucrose 
synthetases are only distantly related. 
Examination of the poly A + mRNA 
from strains homozygous for the sh m 
alleles revealed the presence of a small 
amount of SA-encoded sucrose synthe- 
tase mRNA. This is apparent in Fig. 
38. RNAs from the sh 6233 and sh™ 5933 
strains were included in this experi- 
ment, the details of which are given 
above. Both strains show a small amount 
of a poly A + RNA that is homologous to 








bz m4 










b2 m4 


Fig. 38. Poly A + mRNAs homologous to Sh cDNA in bz m4 , sh m6233 , sh" 15933 , and Sh strains. Poly 
A + RNA purified from immature kernels was fractionated on an agarose gel, transferred to 
nitrocellulose filter paper, and hybridized with cloned Sh cDNA labeled with 32 P by nick-translation. 
Approximately equal amounts of RNA from mutant strains were used and the amount of Sh RNA 
used was roughly eightfold less than the amount of mutant RNA. The three lanes to the right repre- 
sent a prolonged exposure to film of the same region of the gel represented in the three lanes to the 

the Sh cDNA probe and co-migrates 
with the Sh-encoded mRNA present in 
endosperm from an Sh strain. Since the 
two cross-hybridizing mRNAs could be 
distinguished unequivocally by their 
ability to form a nuclease SI -resistant 
hybrid with Sh cDNA, this assay was 
used to titrate the amount of Sh- 
encoded mRNA present in the sh m 
strains. The experiment was the same 
as that described in Fig. 39, except 
that mRNAs from the sh m strains were 
included. The results are shown in Fig. 
40. All three sh m strains contain a 
small amount of Sh -encoded poly A + 
RNA. Compared with the amount pres- 

ent in an Sh strain, the amounts of Sh- 
encoded poly A + RNA in the sh m 
strains varied from slightly more than 
0.5% in sh m6233 and sh™ 5933 to slightly 
less than 1.5% in sh m6258 . Although the 
length of the residual SA-encoded 
mRNA has not been determined for the 
s f l m6258 strain, it appears to be the 
same as that in the wild-type in the 
sh m6233 and sh m5933 str ains (Fig. 38). 

Thus, the Ds mutants of the Sh locus 
appear to produce a small amount of 
S7*-encoded sucrose synthetase mRNA. 
Whether the low level of this mRNA 
results from inhibition of Sh-mRNA 
synthesis or from the rescue of a small 









bz m4 
2.5 12.5 


Fig. 39. Nuclease Sl-resistant hybrids formed between the 500-nucleotide Pst I fragment of the 
cloned Sh cDNA and Sh mRNA. The indicated amounts of Sh and bz m4 poly A + mRNA were 
hybridized with the large Pst I fragment of the cloned Sh cDNA, labeled by nick-translation with 
32 P. Hybridization was done under conditions preventing reassociation of the DNA. Following SI 
digestion and denaturation, the Sl-resistant material was fractionated on a nondenaturing 6% 
polyacrylamide gel. The arrow marks the position of the fragment protected from digestion with SI 
by the S/i-encoded sucrose synthetase mRNA. It is shorter than the undigested fragment by the 
length of the homopolymer tail added during cloning. 

amount of a normal transcript from an 
unstable precursor cannot be deter- 
mined from these data. 

Ds insertion mutations are stable 
in the absence of the controlling ele- 
ment Ac, reverting somatically in its 
presence. To determine whether rever- 
sion is mediated by an increase in 
mRNA level, poly A + mRNA from 
s frm6233 containing Ac was included in 
the experiment shown in Fig. 40 (sh m6233 
+ Ac). It is evident that there is 
substantially more S/i-encoded RNA in 
this strain than in a similar strain lack- 
ing Ac (Fig. 40, sh m6233 - Ac). In the 
mRNA preparation used in this experi- 

ment, the amount of S/i-encoded 
mRNA is about 10% of the amount ob- 
served in an Sh strain. This is presum- 
ably due to the presence of sectors of 
endosperm tissue that have reverted to 
the Sh phenotype. Thus Ac-caused 
somatic reversion of this sh m allele ap- 
pears to be mediated by an increase in 
the level of S/i-encoded sucrose syn- 
thetase mRNA. 

Prior to the analysis of the Sh locus 
in the sh m strains having Ds at the 
locus, the DNA sequence in the im- 
mediate vicinity of the Sh locus was ex- 
amined in several strains showing a 
normal Sh phenotype. The results ob- 




co o co 

m co cn< 

(N <J> tP 

<£> in i 1 <o i r 



CO < 

CO z 


01 >1 


i 1 

< £> co Sh 

-^-S1 f m S .03 lAjg 

1 2 3 4 5 6 7 8 9 10 11 12 

Fig. 40. Nuclease Sl-resistant hybrids formed between the 500-nucleotide Pst I fragment of the 
cloned Sh cDNA and homologous sequences in poly A + mRNA from Sh and sh m strains. The experi- 
ment was carried out as described in the legend to Figure 39. The amount of sh m RNA used in lanes 
(1-3) and (5-6) was 7.5 /*g, while the amount used in lane (4) was 12.5 /xg. The amount of Sh RNAs 
used, from left to right, were 0.01, 0.03, 0.1, 0.3, and 1 \ig. The amount of sh m poly A + mRNA used in 
the experiment displayed at the right was 7.5 ^g and the amounts of Sh mRNA were 0.01 and 0.03 
/ig, as indicated. 

tained in studies of three strains will be 
described here. These include two 
strains derived from stocks obtained 
from B. McClintock and are distin- 
guished by the presence of either the / 
(/ DNA) or the C allele (C DNA) of the 
C locus five -recombination units distal 
to the Sh locus on the short arm of 
chromosome 9. The third strain is the 
W23 X K55 hybrid from Dr. E. Coe at 
the University of Missouri. 

To construct a map of the Pst I, 
Hind III, and Bgl II fragments homolo- 
gous to the cDNA probe, DNA isolated 
from these strains was singly and 
doubly digested with these enzymes, 
fractionated on an agarose gel, trans- 
ferred to nitrocellulose paper and hy- 
bridized to the nick-translated cloned 

cDNA. An autoradiogram of this blot 
is shown in Fig. 41. A restriction map 
that incorporates all of the restriction 
fragments with homology to the probe 
could not be constructed. Such a map 
could be constructed, however, if the 
assumption was made that the frag- 
ments hybridizing to the cDNA probe 
in DNA from the bz m4 strain corre- 
spond to the somewhat homologous 
duplicate sucrose synthetase gene (Fig. 
36). This map is shown in Fig. 42. Since 
the Pst I/Bgl II digests obliterate the 
two large Bgl II fragments, it can be 
inferred that two Bgl II sites lie out- 
side the Pst I sites. The other 1.1 -kb 
Bgl II fragment, seen in Figs. 36 and 
41, is presumed to reside in the dupli- 
cate gene. The 2.0-kb Pst I fragment is 



not detected in the Pst I/Hind III di- 
gests, and places the 0.68-kb Hind III 
fragment within the Pst I fragment. 
Analysis of DNA from two sh m strains 
shows that the appearance of a large 
Hind III fragment correlates with the 
loss of the 4.9-kb fragment from the 
progenitor ISh DNA. This observation 
suggests that the 4.9-kb fragment is 
derived from the Sh locus and that the 
5.1-kb fragment, which persists in the 
mutant DNA, is from the duplicate 
gene. The 0.75-kb Pst I fragment ex- 
tends rightward from the Pst I site in 
the cDNA, since it is not altered in the 
Pst I/Hind III digests. In addition, it 
is shortened in the Pst I/Bgl II digests 
by about 20 bp, which corresponds ex- 
actly to the Pst I-Bgl II distance in the 
cDNA insert. Sst I-digested DNA 
yields a single 2.1-kb fragment and 

places the Sst I sites close to, but out- 
side of, the cDNA insert. The 3' end of 
the gene was tentatively assigned from 
the observation that a 250-bp cDNA 
clone (kindly provided by Dr. B. Burr) 
hybridizes only to the 0.68-kb Hind III 
and 6.2-kb Bgl II fragments. The pres- 
ence of two Hind III fragments cou- 
pled with the fact that the cDNA in- 
sert does not contain a Hind III site in- 
dicates that the DNA represented in 
our clone is not contiguous in genomic 
DNA. Thus it is possible that the 
cDNA clone spans an intervening se- 
quence. The restriction mapping stud- 
ies suggest that the size of the inter- 
vening sequence is 0.2-0.5 kb. 

Examination of the Hind III, Pst I, 
and Bgl II digests reveals a consider- 
able amount of restriction enzyme site 
polymorphism in and around the Sh 

I C W23xK55 

P p /B B H/b H H/p P % B ^B H % P P /B B H /B H H /P 

kb kb kb 

62 s 





t> * 



















68-^ — ^| PHI 

Fig. 41. Autoradiogram of Sh maize DNA digests. The source of each DNA (I, C, W23 X K55) is 
indicated above each set of digests. The enzyme(s) used for each digest is indicated above the ap- 
propriate lane. P = Pst I, B = Bgl II, and H = Hind III. 


X Be 


S258A 5933 6233 

3 '_ 5 < cDNA 

J U Ul L 

W 2 3 x K 55 


E U 3g Si 


Fig. 42. Restriction maps of genomic S/j DNA from three strains. The heavy lines represent 
genomic DNA sequences in and around the Sh locus. The strains from which the DNA was purified 
are indicated above and to the left of each map. DNA represented in the cDNA clone is indicated 
below each map; polarity is shown only in the top drawing. The approximate locations of the three 
Ds sh m alleles is indicated by the triangles above the I map. Restriction enzyme abbreviations: B = 
Bgl II, Be = Bel I, H = Hind III, P = Pst I, S = Sst I, X = Xba I. 

locus (Fig. 41). Most striking are the 
differences in the lengths of the Hind 
III and Pst I fragments between C and 
/ DNA. (The faint 4.9-kb fragment in 
the C DNA probably represents con- 
tamination from genetic recombination 
of sequences normally associated with 
the / DNA.) Polymorphism within the 
transcribed portion of the Sh locus is 
suggested by the Hind III data. The 
estimated length of the sucrose syn- 
thetase mRNA is 3000 nucleotides. 
Since the 2.2-kb Hind III fragment in 
the C DNA is substantially smaller 
than Sh mRNA, the Hind III site 5' to 
the cDNA probably resides in the tran- 
scribed region. Its absence in the / 
DNA indicates that polymorphism has 
occurred within the Sh gene. Restric- 
tion site polymorphism has also been 
observed for Eco RI, Xba I, Bam HI, 
and Bel I. The 2.1-kb Sst I fragment, 
which is either wholly or almost wholly 
within the transcribed region, is iden- 
tical in all of the strains examined so 

DNA derived from strains carrying 
the sh m alleles has also been analyzed 

(Table 2). Interpretation of data ob- 
tained is made difficult by both the ex- 
tensive polymorphism and the presence 
of a duplicate gene. Thus, wild-type Sh 
DNA controls that are of the same ge- 

TABLE 2. Restriction Fragments with 
Homology to Sh cDNA* 



















sh m5933 






s f l m6258A 






sh m6233 






*The lengths (in kb) are given of the major 
fragments hybridizing with Sh cDNA in DNA 
from the indicated strains digested with various 
restriction enzymes. 

tNT = not tested. 



netic background as the mutation be- 
ing studied are always included. 

Southern blots of DNA from one Ds- 
containing sh allele, sh m6258 , indicate 
that this mutation carries an insertion 
of at least 4-6 kb. The Sst I hybridiza- 
tion pattern obtained from DNA 
isolated from an sh m6258 /Sh (W23 X 
K55 background) strain reveals a new 
6.2-kb fragment with homology to the 
probe. In addition, a novel 12.0-kb Bgl 
II fragment has replaced the 6.2-kb pro- 
genitor fragment. Alterations in the 
hybridization patterns obtained from 
two different digests argue that the 
s f l m6258 allele i s the product of an inser- 
tion between the 3' Sst I site and the 
Bgl II site contained in the cDNA in- 
sert. A new 13.0-kb Hind III fragment, 
which probably is derived from the 
4.9-kb Hind III fragment, is evident 
and supports these conclusions. Since 
the cDNA probe spans the site of inser- 
tion, the insertion is probably devoid of 
Sst I, Bgl II, and Hind III sites. Other- 
wise, two new restriction fragments 
should have been observed in at least 
one of the digests. 

The mutable sh allele, sh m5933 , may 
contain an 8-kb insertion. A novel 
13.0-kb Hind III restriction fragment 
has replaced the 4.9-kb progenitor frag- 
ment. Bgl II digestion of the mutant 
DNA yields an 11.0-kb fragment in the 
heterozygote and a concomitant loss of 
the 3.6-kb control / Sh DNA fragment. 
No change in the Sst I hybridization 
pattern was observed. These data in- 
dicate that the sh" 15933 allele contains a 
rearrangement between the Sst I and 
Hind III sites that are 5' to the insert 
DNA. Both the Bgl II and the Hind III 
data yield a net increase of roughly 8 
kb, suggesting the rearrangement is an 
insertion. Analysis of DNA from an 
sh" 15933 homozygote confirms these re- 
sults. In addition, the 10.5-kb progeni- 
tor Xba I fragment has been replaced 
by a 14.6-kb fragment, providing addi- 
tional evidence for an insertion. 

Digestion with Eco RI of DNA 
purified from a strain homozygous for 

another Ds sh allele, sh m6233 , gives rise 
to a 29-30 kb fragment, as compared 
to the 24-kb fragment detected in ISh 
DNA. Xba I digests show that the 
10.5-kb control fragment has been 
replaced by a 13.7-kb fragment. The 
Bgl II, Hind III, and Sst I hybridiza- 
tion patterns were unaltered. These 
data demonstrate that s h m6233 DNA 
contains a DNA rearrangement be- 
tween the Bgl II and Xba I sites that 
are 5' to the cloned DNA. 

In summary, evidence has been ob- 
tained that there are chromosomal 
rearrangements in three independent 
Ds mutations at the Sh locus. These 
rearrangements affect different re- 
gions within the locus. One appears to 
be an insertion within the transcribed 
region of the locus and may well be 
within an intervening sequence. Since 
the other two rearrangements are out- 
side of the sequence covered by the 
Sh cDNA, it cannot be determined 
whether they are insertions or other 
types of rearrangements. Since the ex- 
tent of the transcribed region has not 
been determined, it is also not known 
whether they are within the transcribed 
region. Given the deduced polarity 
of transcription, both of these rear- 
rangements (sh m6233 and sh m5933 ) ap- 
pear to be on the 5' side of the first one 
(sh m6258 ). In the absence of Ac, 
transcripts of the Sh locus have been 
detected in all three strains at levels 
between 0.5 and 1.5% of that present 
in a wild-type strain. The size of the 
transcript has been determined in the 
sh m6233 and sh" 15933 strains and appears 
to be the same as that in a wild-type 
strain. Thus, introduction of Ds re- 
duces, but does not altogether preclude 
expression of the Sh locus. It is not 
known whether the low level of expres- 
sion is consequent on inhibition of 
transcription. The observation that 
there is an insertion within a transcribed 
region, possibly in an intervening se- 
quence, suggests the intriguing pos- 
sibility that the insertion reduces the 
efficiency with which the precursor 



RNA is processed. Finally, the somatic 
reversion of the sh m6233 strain, upon in- 
troduction of the trans-acting control- 

ling element Ac, appears to be medi- 
ated by an increase in the level of Sh- 
encoded sucrose synthetase mRNA. 


L. Mets 

The chloroplast genome, with a mo- 
lecular size of 120-190 kilobase pairs, 
has been a fruitful object of physical 
mapping studies over the past five 
years. The coding sequences for the 
chloroplast rRNAs, numerous tRNAs, 
and several polypeptide gene products 
have been localized in a number of 
species, and many of these genes have 
been sequenced. At the same time, 
techniques for inducing mutations in 
non-Mendelian (presumably chloro- 
plast) genes have been refined in 
Chlamydomonas reinhardii and Oeno- 
thera, and collections of mutations 
with a wide variety of different 
phenotypes are accumulating. Each of 
these approaches to chloroplast gene 
identification and analysis has its 
limitations, however. Many of the mu- 
tants are deficient in one or more 
chloroplast polypeptides, and in some 
cases, mutations at clearly distinct 
genetic loci have indistinguishable 
polypeptide deficiencies. For these mu- 
tants, identification of the primary le- 
sion by biochemical analysis will be 
quite difficult. On the other hand, 
purely physical mapping cannot pro- 
vide the information on chloroplast 
gene function obtainable by careful 
analysis of mutants. To overcome 
these limitations, I am attempting to 
establish a correlation between the 
physical map of chloroplast DNA 
(cpDNA) in C. reinhardii and the ge- 
netic maps of the presumed chloroplast 
genes produced in this alga. Such a cor- 
relation would allow the physical isola- 
tion of chloroplast genes based upon 
their function, and the characterization 

of primary lesions in mutants at the 
level of the altered DNA sequence. 

My first approach to this problem 
(with the collaboration of Ruth Gallo- 
way and Lois Geist) was to develop a 
genetic mapping method in which 
physical site markers on cpDNA could 
be scored in recombination with 
genetic markers. The result would be a 
genetic map which contained sites on 
the physical map. By analyzing the 
meiotic progeny of natural zygotes, we 
found two phenomena that reduce the 
utility of this approach. In zygotes 
that receive heritable contributions of 
chloroplast genes from both parents 
(biparental, or BP zygotes), progeny 
cells carrying recombinant combina- 
tions of the parental genes show ap- 
parently stronger linkages among al- 
leles contributed by the mating-type 
plus (mt+) parent than among those 
contributed by the mating-type minus 
(mi-) parent. This could occur if the 
genome from the mt— parent were 
fragmented either before or during the 
recombination process. A second phe- 
nomenon is that all of the genetic 
markers we have studied show an alle- 
lic bias favoring the inheritance of one 
or the other of the alleles by progeny 
which are genetically recombinant. We 
have not yet determined conclusively 
whether this bias arises from a growth 
advantage of one allelotype over the 
other or from a bias produced in the 
recombination process itself. Both of 
these phenomena tend to obscure 
linkage relationships that might be 
detected by this mapping method. We 
predict that linkage will be detectable 



over only short physical distances. 
Thus, in order to exploit this approach, 
many more genetic markers must be 
isolated. It may also help to study 
recombination in vegetative zygotes or 
somatic fusion products. 

DNA-mediated transformation is a 
rather direct method for determining 
the physical site of mutant genes. The 
results of our genetic mapping efforts 
have encouraged me to attempt trans- 
formation of chloroplast genes in C. 
reinhardii as an alternative. Chloro- 
plast transformation is a special prob- 
lem since the transforming DNA must 
cross not only the plasma membrane 
but also the double-membrane enve- 
lope of the chloroplast. This envelope is 
not permeated by pores, as is the 
nuclear envelope, and it does not trans- 
port any nucleic acids, as far as is 
known. Once inside the chloroplast, the 
donor DNA must become incorporated 
into the recipient cpDNA, preferably 
by homologous recombination. To 
achieve transformation in this context, 
I decided to attempt microinjection of 
a donor cpDNA fragment directly into 
the recipient cell chloroplast. Work has 
progresed in three areas: (1) isolation of 
a mutation in a known chloroplast gene 
and molecular cloning of both the mu- 
tant and wild-type versions of the gene, 
(2) development of methods for physi- 
cally handling cells both during and 
after microinjection, and (3) develop- 
ment of an assay to determine if cells 
have the enzymatic machinery neces- 
sary for performing homologous recom- 

When assaying for transformation, it 
is useful to employ a genetic marker 
which can be clearly selected, which 
has a low spontaneous mutation rate, 
and whose physical location on cpDNA 
is already known. We had previously 
reported a mutant of C. reinhardii 
(10-6C) that synthesizes a structurally 
normal, but enzymatically inactive, ri- 
bulose biphosphate carboxylase 
(Spreitzer et al., Nature, 285, 114-115, 
1980). The large subunit (LS) of the 

protein has an altered isoelectric point, 
and the mutation is inherited in the 
pattern predicted for a chloroplast 
gene. The mutant requires acetate for 
growth (since it cannot photosynthe- 
size) but has intact photosynthetic 
electron transport and photophospho- 
rylation. The acetate requirement and 
the altered LS isoelectric point are co- 
inherited during chloroplast gene re- 
combination. These properties are con- 
sistent with 10-6C being a missense 
point mutant in the LS structural gene, 
which occupies a known position on the 
physical map of C. reinhardii cpDNA 
(Malnoe et al, J. Mol Biol, 133, 
417-434, 1979). Although revertants 
to photoautotrophy have been re- 
covered (R. Spreitzer, personal com- 
munication), their frequency is ex- 
tremely low— much less than one per 
10 8 cells plated. I have isolated 
molecular clones (in pBR322/E. coli 
HB101) of the intact LS gene from 
both wild-type cells and mutant 10-6C. 
These are being sequenced in collabora- 
tion with M. Dron and J. Rochaix to 
identify the site of the mutation. The 
cloned wild-type LS gene will be used 
as donor DNA in an attempt to trans- 
form 10-6C to photoautotrophy. Be- 
cause the cloned segment carries the 
entire LS gene, there is a chance that it 
will be expressed to allow transient 
C0 2 fixation, even if the gene does not 
become stably heritable. 

Despite its small size (6-10 ^m 
diameter), C. reinhardii is probably a 
good candidate for microinjection into 
the chloroplast. It contains a single 
chloroplast, which composes about 
40% of the cell volume and occupies a 
fixed position, essentially filling the 
posterior end of the cell. Thus, it 
should be readily accessible to microin- 
jection, provided the cell is held sta- 
tionary in the proper orientation. The 
following procedure is effective for 
holding the cells. A sterile aqueous so- 
lution of 0.1% poly-DL-lysine is used to 
flood the bottom of a polystyrene petri 
dish. The solution is then poured out 



and the dish air-dried. A drop of cells in 
culture medium is dropped onto the 
dry plate from a height of 60 cm. The 
plate is flooded with culture medium 
and left at room temperature for two 
hours. The cells that cannot be rinsed 
off the plate with several washes of 
medium remain adherent for several 
hours more and can be impaled with 
micropipets. Because many cells are all 
fixed in a plane, numerous cells can be 
impaled relatively quickly. Following 
microinjection, the cells can be embed- 
ded in medium solidified with 0.5% 
agarose so that each cell and its clonal 
descendants can be relocated. Viability 
through the procedure approaches 

The transformation procedure that I 
am attempting will function best if the 
donor DNA is incorporated into the 
recipient cpDNA molecule via homolo- 
gous recombination to replace the mu- 
tant gene with the wild-type version. 
The question arises whether the cells 
have the necessary enzymatic ma- 
chinery for performing homologous 
recombination of cpDNA. The unique 
structure of cpDNA allows an assay 
that may answer this question. cpDNA 
is circular and, in most organisms, in- 
cluding C. reinhardii, includes a large 
segment (containing the rRNA and 
some tRNA genes) that is repeated in 
inverted orientation. Thus, the repeats 

separate two unique regions {A and B) 
whose orientation relative to one 
another has not been established. In 
fact, it is possible that intramolecular 
homologous recombination between 
the repeat regions could occur, invert- 
ing the relative orientation of A and 
B. If intramolecular recombination 
does not occur, then we would expect 
that rounds of cpDNA replication 
followed by mitotic segregation would 
yield clones of cells containing only one 
relative orientation or the other. It is 
also possible that one relative orienta- 
tion would be preferred for functional 
reasons, even if the cells contain en- 
zymes for homologous recombination. 
To determine the relative orientation 
of the A and B regions in a sample of 
cpDNA, I have constructed a recombi- 
nant DNA molecule linking one end of 
A to one end of B. Heteroduplexes be- 
tween this probe and cpDNA should 
produce molecules whose appearance 
in the electron microscope will clearly 
reveal the relative orientations of A 
and B. This work is being done with 
the collaboration of M. Wu. cpDNA 
molecules from various clonal descend- 
ants of individual zygotes will be studied. 
This will establish whether homolo- 
gous recombination can occur in vege- 
tative cells or if one relative orientation 
is preferred. 


M. Collins, R. Karess, R. Levis, C. Murphy, and G. M. Rubin 

The overall aim of our laboratory is 
to determine certain aspects of the 
structure and function of transposable 
genetic elements in Drosophila melano- 
gaster. Transposable elements are DNA 
sequences that, as discrete genetic and 
physical entities, can insert into or ex- 
cise from the genome by mechanisms 

independent of normal homology-de- 
pendent recombination. The insertion 
of a transposable element can affect the 
expression of genes in its vicinity by a 
variety of mechanisms. Transposable 
elements appear to play a major role in 
genetic phenomena; for example, recent 
studies have shown that a large frac- 



tion of spontaneous mutations is the 
result of DNA insertions. Prior to this 
year, our efforts were focused on a class 
of transposable elements of which the 
copia element is the prototype. (For a 
review of this work see Rubin et at, 
Cold Spring Harbor Symp. Quant. 
Biol, 45, 619-628, 1981.) These studies 
were primarily biochemical. It soon be- 
came clear that to understand how 
transposable elements function we 
would have to use genetic as well as bio- 
chemical approaches. 

This year, we have begun to bridge 
this gap by isolating DNA sequences 
from the white locus. The white muta- 
tion was the first mutant reported in 
Drosophila, probably because of the 
dramatic phenotypic change which it 
produces. Null mutants are fully viable 
and fertile, but have white eyes rather 
than the normal brick-red color. Other 
alleles of the locus give intermediate 
colors, or, in some cases, an altered pig- 
ment pattern. Since deposition of both 
major eye pigment groups is reduced, 
the mutation probably does not simply 
block the biochemical pathway by 
which the pigments are synthesized. 
(For a review, see Judd in The Genetics 
and Biology of Drosophila, Volume lb, 
767-800, 1976.) 

There are a number of mutations of 
the white locus that exhibit genetic 
properties analogous to those of estab- 
lished transposable element-induced 
mutations in prokaryotes, yeast, and 
maize. (See Green in The Genetics and 
Biology of Drosophila, Volume lb, 
929-946, 1976.) These properties in- 
clude reversion to wild-type phenotype 
at high rates, and catalysis of chromo- 
somal rearrangements. We report here 
the preliminary molecular analysis of 
three such unstable white mutations, 
w\ w c , and w DZL . We also describe pre- 
liminary molecular studies of a genetic 
syndrome called hybrid dysgenesis. 
Hybrid dysgenesis appears to result 
from the action of a family of trans- 
posable elements whose transposition 
rate is under strict genetic control. 

Molecular Cloning of the 

Sequences of the White Locus of 

D. melanogaster 

R. Levis and G. M. Rubin, in collaboration with 
P. Bingham 

Because neither the RNA nor the 
protein product produced by the white 
locus is known, we employed an indi- 
rect strategy to clone the DNA seg- 
ment containing it. Gehring and Paro 
(Cell, 19, 897, 1980) noted that a cloned 
DNA segment containing the repeti- 
tive, transposable, copia sequence ele- 
ment hybridized in situ to 3C (the desig- 
nation for the polytene chromosome 
region in which the white locus is known 
to be located) in strains carrying the 
apricot allele of white (w a ) but not in 
other strains tested. This suggested 
that a copy of copia was present within 
or in the vicinity of the white locus in 
w a . By isolating this copy of copia to- 
gether with adjacent unique sequences, 
using recombinant DNA techniques, 
we could obtain a unique sequence 
probe from the region of the white lo- 

Using a fragment from a previously 
cloned copy of copia as a probe, we first 
purified a collection of 190 copia-ho- 
mologous lambda hybrid phage from a 
genomic sequence library we had con- 
structed using DNA from Drosophila 
carrying the w a allele. Since there are 
approximately 50 copies of copia in the 
genome, it was probable that the copy 
of copia from the white locus region 
was a part of this collection. In order to 
distinguish which hybrid phage con- 
tained this particular copy of copia, 
DNA prepared from these phages was 
labeled with tritium in vitro and hy- 
bridized in situ to polytene chromo- 
somes from the species Drosophila sim- 
ulans. The polytene chromosome arms 
of D. simulans, a sibling species to D. 
melanogaster, were labeled by copia 
probes at only three sites, none of them 
near 3C. The average length of the Dro- 
sophila DNA insert in these hybrid 
phages was three times the length of a 



copia element. Therefore, each cloned 
DNA segment should have hybridized 
at the site(s) homologous to the se- 
quences adjacent to the copia in the 
cloned segment as well as to the three 
copza-homologous sites. Among the 
first 40 phages tested, one, designated 
Xw a 5.9, labeled 3C in addition to the 
sites of copia labeling. 

A restriction map of the Drosophila 
insert of this phage is shown in Figure 
43, with the position of the copia ele- 
ment indicated. A 3.1-kb Bam HI frag- 
ment, which lies approximately 2 kb 
away from the end of the copia ele- 
ment, was purified as a representative 
of the adjacent sequences. It hybrid- 
ized in situ to a single site in the region 
of the white locus at 3C in the polytene 
chromosomes of both D. simulans and 
D. melanogaster. A more precise locali- 
zation of the unique sequences from 
\w a 5.9 was achieved by hybridizing 
the 3.1-kb Bam HI fragment probe in 
situ to polytene chromosomes from a 
series of strains carrying chromosomal 
rearrangements with breakpoints be- 
tween the white locus and the adjacent 
complementation groups on either 
side. In all cases the hybridization oc- 
curred at the position(s) predicted for a 
probe from within the region bracketed 
by these breakpoints. Therefore the 
copia and adjacent sequences cloned in 
\w a 5.9 must have originated from the 
white locus region. 

In order to isolate white-locus se- 

quences from a wild-type strain, we 
used the 3.1-kb Bam HI fragment of 
\w a 5.9 as a hybridization probe to 
screen hybrid phage from a library con- 
taining segments of DNA derived from 
the Canton-S wild-type strain of D. mel- 
anogaster. (This library was provided 
by J. Lauer and T. Maniatis.) Restric- 
tion maps were made from the DNA of 
seven hybrid phage clones that hybrid- 
ized strongly. The maps could be ar- 
ranged in a single overlapping array, as 
shown in Fig. 44. These cloned seg- 
ments thus originated from a single chro- 
mosomal region of 24 kb. Aside from a 
few restriction-site polymorphisms, 
which might be expected because of 
different genetic backgrounds, the map 
of \w a 5.9 differs from that of the 
Canton-S wild-type only by the inser- 
tion of a copia element at the position 
shown in Fig. 44. 

In order to determine the limits and 
orientation of the chromosomal region 
spanned by this 24-kb cloned DNA seg- 
ment, DNA of individual phages was hy- 
bridized in situ to polytene chromo- 
somes of strains carrying chromosomal 
rearrangement breakpoints between 
the white locus and the adjacent com- 
plementation groups. DNA from phage 
ml. 2 labeled both sides of breakpoints 
between white and zw9, the distal 
neighbor of white, while DNA from 
phage ml.l labeled both ends of break- 
points proximal to white. These results 
indicate that the cloned segment has 

te ' 














■ i 





Fig. 43. A restriction map of the Drosophila melanogaster DNA segment in the \w a 5.9 hybrid 
phage clone. The stippled bar indicates the copia element in this cloned segment; the central solid bar 
indicates the 3.1-kb Bam HI fragment employed in subsequent analyses. 















1 1 





1 1 1 









1 1 






















c c 











_ o 

O A 

































E o o 














1 1 





1 1 1 














1 1 


■o •» 











1 1 

? o 



_— o 













.= o 

o o-o 






X X 









X X 


CO cox 









' ' 

__) X 



i X m 



1 1 m 



I \ m 

9 1 A 

m 6 i 


i X m 




X m 





1 > 




Fig. 44. Restriction maps of the white-locus chromosomal region of Oregon-R (Ore-R) and 
Canton-S (CS) wild-type stocks. The DNA segments contained in the various hybrid lambda phages 
described in the text are indicated below the map. 

the telomere-centromere orientation 
shown in Fig. 44 and that the 24-kb seg- 
ment extends beyond rearrangement 
breakpoints on either side of the white 
locus. Subsequent in situ hybridiza- 
tions using fragments purified from 
these phages have narrowed the maxi- 
mum limits of the locus to a region of 
about 12 kb. 

We have carried out a series of South- 
ern blot hybridization experiments to 
demonstrate that the cloned DNA rep- 
resents the genomic arrangement of 
these sequences. Co-migration of homol- 
ogous cloned and genomic DNA frag- 
ments spanning the cloned region 
showed, within the resolution of the 
technique, that there were no deletions, 
insertions, or substitutions during the 
cloning of these segments. We have 
also constructed a restriction map of 
the corresponding segment from an 
Oregon-R wild-type stock, shown in 
Fig. 44, using similar genomic South- 
ern blots in conjunction with maps of 
cloned segments from the w DZL allele, 

which was derived from an Oregon-R 
wild-type. Several restriction site dif- 
ferences between the two maps indicate 
that there are some sequence differ- 
ences between these two wild-type iso- 
lates. However, there do not appear to 
have been any large (>200 bp) inser- 
tions or deletions within the 24-kb re- 
gion. Genomic Southern blot experi- 
ments also suggest that there are no 
large blocks of sequences within this 
region repeated elsewhere in the 
genome. In situ hybridizations using 
the hybrid phages are consistent with 
this conclusion. 

Molecular Analysis of the w dzl 

R. Levis and G. M. Rubin 


White dominant zeste-like {w"*^) is a 
spontaneous mutation recovered by 
Bingham (Genetics, 95, 341, 1980) in a 
wild-type Oregon-R stock of D. melano- 
gaster. The w DZL mutation is dominant 



and exhibits paradoxical dosage ef- 
fects; males are nearly wild-type while 
both homozygous and heterozygous fe- 
males have a yellow eye color. Bingham 
considered the mutation to be an allele 
of the white locus on the basis of recom- 
bination mapping, which placed it to 
the proximal side of all known alleles 
but distal to roughest, the next most 
proximal complementation group. The 
mutation is unusual in that it reduces 
the expression of the white locus in cis 
or trans only when the two homologues 
carrying the white locus are able to 
synapse in the white-locus region. 

Our primary interest in the w DZL mu- 
tation stems from its instability; about 
0.1% of the progeny of w DZL flies have 
different alleles of the white locus (Bing- 
ham, Cold Spring Harbor Symp. Quant. 
Biol., 45, 519-525, 1981). These include 
derivatives that are phenotypically wild- 
type and others that are white-eyed. 
While the original mutation as well as 
many of its derivatives have no chro- 
mosomal abnormalities detectable by 
examination of polytene chromosomes, 
some of the derivatives contain dele- 
tions or inversions that in all cases 
have one breakpoint in the region of the 
white locus. The generation of chromo- 
somal rearrangements by w DZL — and 
its instability— suggest that this allele 
results from a transposable element in- 
sertion at the white locus. To test this 
hypothesis, we isolated, by recombi- 
nant DNA cloning, segments of the 
DNA from the white locus in a strain 
carrying the w DZL allele and compared 
them to those of the wild-type. 

A hybrid lambda phage library of 
genomic DNA segments from w DZL 
was constructed and screened with 
several probes from the 24-kb white- 
locus region described above. Restric- 
tion maps were prepared from the DNA 
of 21 phages recovered in this screening. 
These maps were arranged in an over- 
lapping fashion to define the map of a 
48-kb continuous segment, which in- 
cludes the entire 24-kb region known to 
contain all alleles of the white locus. 

The only difference we could recognize 
between the map of this region in w DZL 
and that of Oregon-R is the presence of 
a 13-kb insertion at the position indi- 
cated in Fig. 45. 

We have begun to investigate whether 
there is a correlation between the pres- 
ence of this insertion in the white-locus 
region and the mutant phenotype by 
examining the DNA of phenotypically 
wild-type revertants of w DZL . Prelimi- 
nary Southern blot analysis suggests 
that several revertants retain a part, 
but not all, of the insertion. The portion 
remaining varies between 2 and 4 kb in 
the three revertants studied. 

The White-ivory Mutation and Its 

R. Karess and G. M. Rubin 

The ivory allele of the white locus (w l ) 
was isolated by Muller in 1920 as a 
spontaneous mutant. Its genetic prop- 
erties are briefly summarized here and 
are reviewed in detail by Green (The 
Genetics and Biology of Drosophila, 
Volume lb, 929-946, 1976). w l has 
been mapped genetically to the left of 
white-cherry, and to the right of the 
white-buff allele. The ivory allele re- 
sults in a faintly tinted eye, somewhat 
darker in homozygous females than in 
males, as it is not dosage compensated. 
The phenotype includes a reduced fre- 
quency of recombination between w l 
and flanking alleles of white. White- 
ivory has a relatively high spontaneous 
rate of reversion to wild-type, which re- 
stores both the original eye color and 
the normal recombination frequency 
within the locus. The frequency of re- 
version is approximately 5 per 10 5 X 
chromosomes in homozygous females, 
and about one-tenth this rate in males 
and deletion-heterozygote females. X 
irradiation results in nearly equal re- 
version rates in both males and fe- 
males. Exchange of outside markers 
does not accompany the reversion of w l 
to wild- type, and conditions known to 



= B 

H H x 

Fig. 45. Restriction map of the white-locus region showing the positions and sizes of the 
cation and w c and w DZL insertions. 


augment the frequency of recombina- 
tional events fail to increase the fre- 
quency of w l reversion, suggesting that 
other mechanisms are involved. On the 
basis of these and other genetic consid- 
erations, it has been proposed that the 
w l phenotype results from a small du- 
plication of a portion of the white locus. 
The w' allele was of interest to us both 
as an unstable arrangement of DNA 
and as the immediate precursor to the 
highly unstable white-crimson allele 
(see below). The molecular cloning of 
the wild- type locus allowed us to ex- 
amine the DNA structure of the mutant 
w' allele. Southern blot analysis of re- 
striction nuclease-digested w l DNA 
demonstrated that w l was indeed a 
small tandem duplication of approxi- 
mately 3 kb within the white locus 
region (Fig. 45). Aside from single re- 
striction-site polymorphisms, no other 
differences between w l and the wild- 
type DNA sequences were detected. 
Several clones containing the w l gene 
were recovered from a recombinant li- 
brary constructed from the lambda vec- 
tor Charon 28 and total w l fly DNA, 

and analysis of these confirmed the 
presence of the 3-kb duplication. Using 
the cloned w i DNA as a probe of a 
whole-genome Southern blot, no restric- 
tion fragments other than those ex- 
pected from the white locus were found, 
suggesting the absence of DNA se- 
quences not originating from the locus. 
The genomic structure of three pheno- 
typic revertants of w l (provided by M. 
M. Green) were examined. Southern 
blot analysis of one x-ray revertant 
and one spontaneous revertant re- 
vealed them to be simple deletions of 
one copy of the duplicated sequences, 
leaving a single copy as in the wild- 
type, thus confirming that the 3-kb du- 
plication is responsible for the ivory 
phenotype. To the limits of resolution 
of these Southern analyses (±100 base 
pairs), these two reversions exactly re- 
stored the wild-type DNA sequence in 
the region. 

The third phenotypic revertant that 
we examined proved to be more com- 
plex than either w l or the wild- type 
allele. Analysis of this isolate revealed 
that the phenotypic reversion to wild- 



type was accompanied by the insertion 
of a DNA sequence of approximately 2 
kb into the site of the ivory lesion. 
Moreover, in addition to this insertion, 
a portion of the original duplicated 3-kb 
sequence was removed, presumably in 
the same event. The inserted 2-kb DNA 
sequence was found to hybridize to ap- 
proximately 50 restriction fragments 
of a whole-genome Southern blot, most 
of which shared several restriction sites 
in common with the inserted element. 
It is hoped that analysis of the genomic 
structure of this and other ivory rever- 
tants will illuminate the range of 
events possible in this small, dupli- 
cated DNA sequence. 

The Molecular Basis of w c 

M. Collins and G. M. Rubin 

White-crimson (w c ) is a highly muta- 
ble allele of the white locus. It was 
derived from w' in a screen for x-ray in- 
duced revertants of w\ and confers a 
light reddish-orange eye color as a 
recessive phenotype. (For a review of 
the genetic properties of w c , see Green, 
The Genetics and Biology of Droso- 
phila, volume lb, 929-946, 1976.) w c 
mutates at a frequency of about 1/1000 
to wild-type, w\ and white-eyed deriva- 
tives. The wild-type revertants are 
mutationally stable and phenotypi- 
cally indistinguishable from wild-type. 
The w l revertants are phenotypically 
identical to the original white-ivory, 
and revert to wild- type at the same low 
frequency characteristic of the original 
white-ivory allele. The white deriva- 
tives fall into two classes: stable and 
unstable. The stable derivatives in- 
clude deletions which have one end- 
point at the white locus, and extend to 
the left or right of the white locus. The 
unstable white derivatives mutate at 
frequencies similar to w c , generating 
wild-type, w l , and w c derivatives. The 
w c derivatives of mutable white alleles 
retain the mutability of the original w c . 

The w c allele is also capable of 
transposition from the X chromosome 
to autosomes. In these new autosomal 
positions, w c retains its property of 
mutability, mutating to white and w d \ 
a phenotype lighter than w c . Genetic 
data suggest that only a portion of the 
white locus has been transposed. 

Genetic evidence indicates that mu- 
tation of w c is not mediated by normal 
homologous recombination events. Mu- 
tations occur in clusters, suggesting a 
premeiotic origin, and are not accom- 
panied by recombination of flanking 
markers. Furthermore, mutations oc- 
cur at similar frequencies in males and 
females, while meiotic recombination is 
virtually absent in male Drosophila. 
Green has proposed that w c is an inser- 
tion mutation, and that it mutates to 
wild-type by excision of the insertion. 
This insertion is hypothesized to be- 
have like prokaryotic insertion ele- 
ments that are capable of excision and 
transposition and that generate dele- 
tions with one endpoint at the site of 

We have begun to analyze the struc- 
ture of the white-crimson allele at the 
DNA level to determine the nature of 
the white-crimson lesion. DNA from 
flies carrying the w c allele was digested 
with various restriction enzymes, size 
fractionated on agarose gels, and trans- 
ferred to nitrocellulose. These Southern 
blots were then probed with wild-type 
white-locus DNA. The restriction pat- 
tern of the white-locus region of white- 
crimson DNA was then compared to 
that of white-ivory and wild-type DNA. 
These experiments revealed that the 
white-crimson allele contains an inser- 
tion of 1 1 kb of DNA into the center of 
the w l duplication, as shown in Fig. 45. 

In order to demonstrate that the 
11-kb insertion is responsible for the w c 
phenotype, wild-type, and w l rever- 
tants oi w c were analyzed by whole 
genome Southern blots to see if rever- 
sion is accompanied by excision of the 
insertion. We have examined six in- 
dependent wild- type revertants of w c , 


and in all six cases reversion is accom- ern blot showing reversion of w c to 

panied by excision of both the 11 -kb in- wild-type is shown in Fig. 46. 

sertion and the 3-kb w* duplication, We have also analyzed the structure 

restoring gene structure to wild-type, of five independent revertants of w c to 

An example of a whole genome South- w l by whole-genome Southern blots. In 

w +J w B w* w c CS 

*m- z m 

W *™ £*"* 

.7 kb 


r- — o fc 


Hmdm/Xba I l 

6.5 kb 2.6kb 





■-. *-"' 







a "° 
E .£ 




























en x 

i i 








1.7 kb 


2.9 kb 


3.5 kb 


Fig. 46. The left panel is a whole-genome Southern blot of Hind III/Xba I-digested DNA from 
w c , Canton-S (CS), and three independent w c revertants to wild-type (w +J , w +B , w +A ). The hybridi- 
zation probe used is indicated by the dark line on the restriction map of the wild-type white locus 
(w + ) drawn below. The fragments detected by this probe for the wild-type white locus are indicated 
below the map. Sizes of fragments were determined by comparison with a Hind III digest of lambda 
DNA. The extra restriction fragment derived from the w c insertion is absent in those revertants that 
generate fragments identical to wild-type DNA. The right panel is a whole-genome Southern blot of 
Bam Hi/Sac I-digested DNA from w l and from two w l derivatives of w c (w' F and w' G . The probe is 
the same one used in the left panel. The restriction map of the w l allele is drawn below, with the 
fragments detected by this probe indicated below the map. 



all five cases, the 11-kb insertion has 
been excised, leaving the w l duplication 
intact. A comparison of gene structure 
in the original white-ivory and two 
white-ivory revertants is also shown in 
Fig. 46. These results demonstrate that 
the w c mutation is due to the insertion 
of an 11-kb piece of DNA into the 
center of the w l duplication. 

Molecular Analysis of Hybrid 

G. M. Rubin, in collaboration with P. Bingham 

Evidence for genetic control of the 
transposition rate of a family of Droso- 
phila transposable elements comes 
from studies of hybrid dysgenesis, a 
"syndrome of correlated genetic traits 
that is spontaneously induced in hy- 
brids between certain mutually inter- 
acting strains, usually in one direction 
only" (Kidwell, Ki dwell, and Sved, 
Genetics, 86, 813, 1977). These traits 
can include sterility, male recombina- 
tion, mutation, and chromosomal aber- 
ration. Drosophila melanogaster exhib- 
its at least two independent systems of 
interacting strains: I-R and P-M. Evi- 
dence reviewed by Engels (Cold Spring 
Harbor Symp. Quant. Biol., 45, 561- 
565, 1981) indicates that the genomes of 

certain strains, called P strains, con- 
tain multiple copies of an apparently 
mobile element known as the P factor. 
A cross between a male from such a P 
strain to a female from a strain which 
lacks P factors, called an M strain, re- 
sults in hybrid dysgenesis. The P fac- 
tors do not produce dysgenesis within 
the P strains, but do so only when 
placed in the cytoplasmic background 
of an M strain. Many of the mutations 
produced by hybrid dysgenesis are un- 
stable, are thought to be insertion mu- 
tations, and may be due to insertion of 
the P factor itself. 

These and other observations re- 
viewed by Engels (op cit.) led to the pro- 
posal that the P factor is a transposable 
element whose rate of transposition is 
under strict genetic control. We have 
recently obtained biochemical evidence 
in support of this hypothesis; two hy- 
brid dysgenesis-induced mutations in 
the white locus isolated by Simmons 
and Lim (Proc. Nat. Acad. Sci, 77, 
6042, 1980) are due to DNA insertions 
of 1.1 and 1.4 kb (see Fig. 47). Southern 
blot and in situ hybridization experi- 
ments revealed that these inserted 
DNAs are members of a family of DNA 
sequences absent in the several M 
strains examined but present as a dis- 
persed repetitive sequence in P strains. 


CO — 

— o o 



\ 12 / 

B~ B 

at " 



a 2 




c o C 





— u — 




c o 





fc — 










o o 














1 I 










1 1 

Fig. 47. Restriction map of the white-locus region showing the positions of two insertion muta- 
tions induced by hybrid dysgenesis. 



We have also examined the structure of 
one revertant of each of these muta- 
tions. In both cases the structure of the 
white locus was returned to wild-type, 
suggesting that reversion is due to pre- 
cise excision of the inserted DNA. We 

are now performing similar analyses of 
several additional hybrid dysgenesis - 
induced white mutations that have 
been isolated by Margaret Kid well of 
Brown University. 


A. Spradling, J. Levine, S. Parks, and B. Wakimoto 

In last year's Report (Year Book 79), 
we described eggshell (or chorion) syn- 
thesis in the fruitfly Drosophila melan- 
ogaster. Our work has focused on the 
properties of chorion protein genes, 
since their temporally regulated ex- 
pression in ovarian follicle cells under- 
lies eggshell production. Eggshell pro- 
tein genes occur in the genome in at 
least two clusters. Genes for two of the 
major proteins (designated s38-l and 
s36-l) are located on the X chromosome 
at polytene band 7F1-2, while the si 8-1 
and si 5-1 chorion proteins are encoded 
at 66D11-15 on the third chromosome. 
Of particular interest has been the find- 
ing that all four of these genes undergo 
amplification as part of normal follicle 
cell differentiation. Consequently, this 
system provides examples of two fun- 
damental developmental mechanisms— 
the replication and the transcription of 
specific genomic components— that are 
particularly accessible to genetic and 
molecular studies. 

If a genetic approach to this system 
is to succeed, however, detailed knowl- 
edge of normal chorion gene function 
will be required. This year several of 
our studies have highlighted the poten- 
tial significance of the dispersed, clus- 
tered arrangement of the genes for the 
four major chorion proteins. Cloning 
and characterization of the genomic 
DNA surrounding these genes have led 
to the discovery of additional closely 
linked genes which are active in follicle 
cells during the time of eggshell pro- 
duction. Possible explanations for this 

extensive clustering are (1) evolution- 
ary—the clustered genes may correspond 
to a multigene family— (2) morpholog- 
ical—they may represent components 
of a specific eggshell subassembly, e.g., 
the pillars or the endochorion roof— 
and (3) functional— they may be sub- 
ject either to some common temporal 
regulation or to coordinate regulation 
of amplification. 

To test the simple idea that the 
genes in a cluster are transcribed coor- 
dinately, we measured the amount of 
each RNA product present in follicle 
cells during oogenesis. No obvious 
relationship was found between the lo- 
cation of a gene and its time of expres- 
sion. Even genes separated by only 1-2 
kb within a cluster may be controlled 
independently, giving rise to RNAs 
that differ in their temporal patterns of 
accumulation. The gene products coded 
within a single cluster also differ sig- 
nificantly in abundance, suggesting 
that rates of accumulation are also de- 
termined individually. However, the 
transcripts characterized from the 
third chromosome are generally smaller 
and expressed later in oogenesis than 
those derived from the X chromosome. 
Clearly, the clustering of subsets of the 
chorion genes has not occurred totally 
at random. 

A much closer relationship appears 
to exist between gene clustering and 
the process of amplification. This year, 
our studies of amplification showed 
that the chorion genes within a cluster 
are amplified equally. All the genes in 



the 7F region of the X chromosome are 
amplified about 16-fold, in contrast to 
the 60-fold replication of the 66D gene 
cluster on chromosome 3. Furthermore, 
amplification depends on the continu- 
ity of the genes in the cluster. When 
the two halves of the chorion gene clus- 
ter of the X chromosome are separated 
as a result of an inversion, amplifica- 
tion occurs in only one of them. 

We have begun to define the mecha- 
nism of amplification by demonstrat- 
ing that the differential replication in- 
volves a large continuous chromosomal 
domain. Each domain consists of a 
maximally amplified central region 
10-20 kb in length and contains the 
chorion genes, as well as flanking se- 
quences which are amplified to a lesser 
extent. As the distance along the chro- 
mosome from the gene cluster in- 
creases, amplification decreases until 
eventually it can no longer be detected. 
We have cloned the entire amplified do- 
main surrounding the chorion genes at 
7F and found that it measures about 90 
kb. The amplified domain surrounding 
the chorion genes on the third chromo- 
some is slightly larger— an estimated 
100 kb. 

The discovery of a new sort of chro- 
mosomal unit— the unit of amplifica- 
tion—is exciting because of its possible 
implications for understanding how 
DNA replication is controlled in eukar- 
yotic organisms. A great deal remains 
to be learned about DNA replication 
during development. Currently, it is 
not known how the S phase is initiated, 
how the onset and rate of replication of 
specific sequences is determined, or 
how a second round of replication is 
prevented during a typical cell cycle. 
Even the existence of specific origins 
of replication in eukaryotic chromo- 
somes such as those found in prokaryo- 
tes and eukaryotic viruses remains in 
doubt. We hope that a detailed analy- 
sis of chorion gene amplification, in 
which replication appears to be con- 
fined to as few as two chromosomal 
sites, will provide insights into the pro- 

cesses that control replication during 

The polytene chromosomes observed 
in the larval salivary gland have long 
served as a useful model of replication 
in polyploid insect cells, such as 
Drosophila follicle cells. In another fly, 
Rhyncosciara, amplification of small 
salivary gland chromosome regions 
takes place near the end of larval devel- 
opment, as indicated by a relative in- 
crease in the staining of several speci- 
fic bands. The occurrence of such 
"DNA puffs" is followed shortly by 
RNA puff formation and leads to the 
production of protein constituents of 
the larval cocoon, as shown by a series 
of studies in Francisco Lara's labora- 
tory (see Cell, 17, 827-833, 1979). The 
many similarities between DNA puff 
formation in Rhyncosciara and Dro- 
sophila chorion gene amplification sug- 
gest that these two processes utilize 
common mechanisms. 

Do polytene chromosome bands there- 
fore correspond to units of replication? 
In the case of chorion gene amplifica- 
tion, one cannot ask this question 
directly, since follicle cells do not them- 
selves contain visible polytene chromo- 
somes. We have determined the region 
of the Drosophila salivary gland chro- 
mosomes that corresponds to the am- 
plified sequences of the X chromosome 
by in situ hybridization (Fig. 51). It 
was found to encompass the two bands 
7F1-2 and 7F3-4 as well as part of the 
faintly banded 7E region. Therefore, 
the domain of X chromosome amplifi- 
cation appears to be larger than re- 
gions forming single bands in the sali- 
vary gland chromosomes. 

Structure of Two Chorion 
Gene Clusters 

A. Spradling and J. Leuine 

Chromosomal segments composed of 
the central 15 kb of the DNA contain- 
ing the chorion gene clusters have 
been subcloned in the plasmid vector 



pBR322 and extensively mapped with 
restriction enzymes. The structure of 
identified chorion genes and the lo- 
cation of regions transcribed into folli- 
cle cell RNA of as yet unknown func- 
tion is shown in Fig. 48. (A new system 
of nomenclature for chorion proteins 
has recently been agreed upon. All pro- 
teins are designated by the letter s [for 
shell], a number indicating the molecu- 
lar weight [in kilodaltons], and a digit 
to identify different proteins within a 

single size class. Thus s38-l replaces 
the previous c38 or c38-60 or C2.) The 
direction of transcription of s38-l, 
s36-l, sl8-l, and sl5-l was determined 
as illustrated in Fig. 49. End-labeled re- 
striction fragments were tested for 
their ability to prime cDNA synthe- 
sis from the complementary chorion 
mRNA. Priming by the 32 P-labeled 
strand led to an increase in the size of 
the primer fragment detectable by gel 
electrophoresis. The results indicated 

Bom HI 
Ave I 

•36-1 (38-1 

5' 3' 3' 3' 

i iii 
i i 

SF9 I6C3 TC8 

i i 
ii till 

I I I 

I I i 

I i 


p 104.60 





EcoRI i 

BemHI • 







tl8-l »I3-I 

3' 3' 3' 3' 




p 302. 77 

p 302. 42 

Fig. 48. The structure of two chorion gene clusters defined by cloned DNAs. The restriction map 
of two genomic fragments (X103 and X104) and subclones derived from them (pl03.48, pl03.29, etc.) 
are shown in (A). These sequences are derived from the 7F1-2 region of the X chromosome. Regions 
complementary to follicle cell poly(A)-containing RNA are indicated by thick lines. The location of 
genes for s38-l and s36-l was determined by hybridization with the corresponding cDNA clones 
(Dmc7C8, Dmcl6C3, and Dmc5F9). (B) shows the restriction map of the third chromosome gene 
cluster at 66D11-15, as deduced from genomic DNAs (X302, p302.77, and p302.42). sl8-l and sl5-l 
coding sequences were located by hybridization to the cDNA clones Dmc5G2 and Dmc7B7. 



S38-1 s36-l sl8-l sl5-l 


12 3 4 5 6 7 8 

Fig. 49. Direction of transcription of four chorion genes. The indicated restriction fragments 
were labeled at their 5' or 3' ends, denatured, and hybridized to RNA extracted from 100 stage 12-14 
egg chambers. Hybridized fragments were then used to prime cDNA synthesis by reverse transcrip- 
tase in the presence of unlabeled nucleoside triphosphates. To determine if the labeled or unlabeled 
strand of each fragment had functioned as a primer, the reactions were analyzed by electrophoresis 
on 8% acrylamide gels. Knowledge of which strand is complementary to mRNA allows the direction 
of transcription to be deduced by reference to the restriction maps of Fig. 48. (1) 0.73-kb Pvu II-Xho 
I fragment of pl04.41 labeled at the Xho I 3' end. (2) 0.58-kb Xho I-Pvu II fragment of pl04.41 la- 
beled at the Xho I 3' end. (3) 0.12-kb Pvu II-Bam fragment of pl03.48 labeled at Bam 3' end. (4) 0.48 
Bam-Pvu II fragment of pl03.48 labeled at Bam 3' end. (5) 0.15-kb Eco RI-Pvu I fragment of 
Dmc5G2 labeled at 5' Eco RI end. (Eco RI sites are located at the junctions of Drosophila and vector 
DNA in the cDNA clones.) (6) 0.46-kb Pvu I-Eco RI end labeled at 5' RI end. (7) 0.24 kb Eco RI-Pvu I 
fragment of Dmc7B7 labeled at Eco RI 5' end. (8) 0.34-kb Pvu I-Eco RI fragment of Dmc7B7 labeled 
at Eco RI 5' end. 

that in both clusters, two major cho- 
rion genes are transcribed in tandem. 

Detailed comparison of chorion 
cDNA' and genomic restriction maps 
provides evidence for the presence of at 
least one intron in the s36-l gene. In 
the coming year, we hope to largely 
complete the structural analysis of the 
region containing the gene clusters. All 
the follicle cell transcripts deriving 
from within these regions will be iden- 
tified. The translational capacity of 
each RNA will be tested to determine 
which ones represent mRNAs for cho- 
rion structural proteins. Finally, tran- 

scription and processing steps which 
lead to production of functional cho- 
rion messenger RNAs will also be 

The Temporal Program of Chorion 
Gene Expression 

S. Parks and A. Spradling 

A minimum requirement for study- 
ing the temporal control of chorion 
gene expression during oogenesis is a 
detailed knowledge of the gene's nor- 
mal developmental program. Only then 



will it be possible to recognize specific 
alterations that may occur in mutant 
strains. At present we are concentrat- 
ing on the expression of those chorion 
genes located in the gene clusters at 7F 
and 66D. Initial studies have involved 
hybridizing labeled restriction frag- 
ments complementary to specific cho- 
rion mRNAs or to particular genomic 
regions to RNA isolated from staged 
egg chambers and immobilized on fil- 

ters following size fractionation by gel 
electrophoresis (Northern blotting). 

Figure 50(A and B) shows the pat- 
terns of developmental expression of 
sequences derived from the X chromo- 
some and third chromosome, respec- 
tively. Both s38-l and s36-l mRNAs 
are present in stages 11-13, but then- 
concentration is maximal during stage 
12. The sl8-l mRNA is expressed dis- 
tinctly later. It is found mainly in 


, O r N O t fOr-NCO^- 


.7 . 


B. 3rd CHROMOSOME (66D11-15) 

OrN O^ 

9 o t- cm n*t 

o> o «- w n t 

?Or CV CO ^ 




Fig. 50. The expression during oogenesis of poly(A)-containing RNAs coded by the X chromo- 
some cluster (A) and the third chromosome cluster (B). Poly(A)-containing RNA was isolated from 
100 egg chambers of stages 1-9, stage 10, stage 11, stage 12, stage 13, or stage 14. Filters containing 
the RNAs were prepared following glyoxal-treatment and separation in 1.5% agarose gels. Each 
filter was hybridized to the indicated 32 P-labeled DNA probe. RNAs complementary to the probes 
were detected by autoradiography. Glyoxal-treated restriction fragments of pBR322 served as size 
markers. The probes were (see Fig. 48) (Al) pl04.60, (A2) Dmc5F9, (A3) Dmcl6C3, (A4) pl03.29; (Bl) 
Dmc5G2, (B2) Dmc7B7, (B3) the 1.15-kb Hind III fragment of p302.77, (B4) p302.42. 



7A B C 

i I ! 

i a i b II I c 


v *.» F 8AB C E 

v : Ml 

536 ,38 

d I e | f | g | In 1 1 i Ljjjkj I I m in | n i 1 I tl I I o I 


5 Vb 

2304 1813 704 104 103 015 1406 2203 

b b m *% k 

ft •*• e H^ 

2 , 


' f ',k • •*§ 

e a 

e a 

e a 

e a 

e 12 






40 kb 

Fig. 51. Amplification of the X-chromosome chorion gene region. A map of Eco RI sites in about 
100 kb of chromosomal DNA surrounding the X-chromosome chorion gene cluster is shown. (Sites 
smaller than 1 kb have been omitted. The order of several unlabeled fragments from the 1904 region 
is still uncertain.) The location of the Drosophila DNA inserts from a series of overlapping clones 
used as probes for these genomic RI fragments is indicated, as well as the RNA complementary 
regions. In situ hybridization of cRNA complementary to 2203 and 2304 to a polytene chromosome, 
indicating the cytogenetic extent of this region, is shown at the top of the figure. Each cloned 
genomic segment was labeled and hybridized to Rl-digested embryo DNA (e) and stage-13 egg- 
chamber DNA (a) which had been electrophoresed on a 0.8% agarose gel and transferred to a 
nitrocellulose filter. Only about one-tenth (1813, 704, 104, 15) or one-fourth (2304, 103, 1406, 2203) as 
much egg-chamber as embryonic DNA was present to help equalize the differences in labeling due to 
amplification. The position on the gels of specific genomic RI fragments is shown by lower-case let- 
ters. All the hybridizations also included DNA complementary to a nonamplified genomic region 
(Dme8D10, see P.N.A.S., 76, 1096-1100, 1980) as an internal control (indicated by z). Below, the ex- 



stage 13, but reduced amounts are also 
present during stage 14. The si 5-1 
gene, which is located about 1 kb from 
the sl8-l gene, displays yet another 
pattern of expression, si 5-1 mRNA is 
confined almost entirely to stage 14, 
the last stage of egg-chamber develop- 
ment. Each mRNA is present only dur- 
ing the known synthetic period of the 
corresponding protein. Thus synthesis 
of these proteins is probably controlled 
by mRNA availability rather than at 
the translational level. 

Both sl8-l and sl5-l probes hy- 
bridize not only to their respective 
mRNAs (0.8 kb and 0.7 kb) but also, 
more weakly, to larger RNAs. These 
larger RNA species could represent 
precursors to the mRNAs or indepen- 
dent transcripts which share some 
sequences recognized by the probes. 

The actual complexity of the RNAs 
coded by these chromosomal regions is 
indicated by experiments in which 
genomic fragments were used to probe 
the RNAs. As many as four discrete 
transcripts complementary to some re- 
striction fragments were detected. The 
origin and function of these RNAs re- 
mains to be determined; a few of them 
may be mRNA precursors. However, 
their unique size and developmental 
specificity can be most simply ex- 
plained by assuming they code for ad- 
ditional chorion proteins. For example, 
the 1.6-kb RNA labeled by the 6.3-kb 
fragment derived from 7F is found 
mainly in stage 11. A follicle-cell 
mRNA (E2) with these properties has 
been described previously (Cell, 15, 
589, 1979) and may code for chorion 
protein s60-l. If each chorion gene 
cluster codes for 4-8 proteins as these 
preliminary results suggest, this would 

account for nearly half of the eggshell 
components so far identified. 

The Mechanism of Chorion Gene 

A. Spradling, J. Levine, and B. Wakimoto 

In last year's Report, we described 
experiments designed to elucidate the 
mechanism of chorion gene amplifica- 
tion. We have extensively pursued one 
powerful approach to this question by 
mapping restriction sites of large chro- 
mosomal regions in both embryonic 
DNA, which contains only unamplified 
chorion gene sequences, and in stage 
13 egg chamber DNA, where amplified 
DNA is present at maximal levels. Any 
differences in the restriction maps of 
amplified and unamplified sequences 
would be of mechanistic significance. 
Furthermore, by comparing the rela- 
tive intensities of individual restriction 
fragments, it is possible to get an idea 
of the extent to which chromosomal 
sequences surrounding the chorion 
genes are amplified in follicle cells. 

The results of these experiments are 
shown in Figs. 51 and 52. Throughout 
the entire chorion gene regions, there 
are no changes in the restriction map of 
embryonic DNA associated with am- 
plification. This observation rules out 
(in their simplest form) models of am- 
plification wherein a discrete chromo- 
some segment is excised and replicated 
extrachromosomally or as a tandem re- 
peat within the chromosome. When the 
extent of amplification of each Eco RI 
restriction fragment in these regions is 
calculated, an additional fact is evident 
(see Figs. 51 and 52). Amplification oc- 
curs in a discrete domain consisting of 

tent of amplification of each Eco RI fragment is shown as a function of position within the cloned 
region. The number of copies/genome of a restriction fragment was calculated as the ratio of its 
specific labeling intensity in stage-13 egg-chamber DNA compared to embryonic DNA. The specific 
labeling intensity of a restriction fragment was its absolute labeling intensity divided by the inten- 
sity of the 2.6-kb fragment complementary to Dme8D10, which served as a control for the total 
amount of genomic DNA present on the filter. 




> S ' .' • 

66 C D E 68A C E 

k , I ,, m ,n M o 

b ,c„d,e, f , g „h, i , 1, K ,',,"',",| 


e a e a e a 

ea ea ea ea 


i — rrii 

Fig. 52. Amplification of the third chromosome chorion gene region. A map of the Eco RI sites in 
about 100 kb of chromosomal DNA surrounding the third chromosome chorion gene cluster and its 
RNA complementary regions is shown. Fragments less than 1 kb were omitted. The location of the 
Drosophila DNA inserts from a series of overlapping clones that were used as probes for these 
genomic Eco RI fragments is indicated. Each cloned genomic segment was labeled and hybridized to 
Eco Rl-digested embryo DNA and stage-13 egg-chamber DNA which had been electrophoresed on a 
0.8% agarose gel and transferred to a nitrocellulose filter. Only about one-tenth (905, 31, 302, 218, 
1003) or one-half (2101, 1704) as much egg-chamber as embryonic DNA was present on the filter to 
help equalize the differences in labeling due to amplification. The position on the gels of specific 
genomic Eco RI fragments is shown by lower case letters. Below, the extent of amplification of RI 
fragments is shown as a function of position within the cloned region. The number of copies/genome 
of a restriction fragment was calculated as in Fig. 51. Normalization was based on the relative 
amounts of DNA transferred to the filters as measured by the diphenylamine reaction. 



a central maximally amplified region 
and two flanking regions in which the 
extent of amplification decreases as 
the distance along the chromosome 
from the central region is increased. 
The location of the central region corre- 
sponds with the position of the tran- 
scribed sequences. The extent of ampli- 
fication of all the sequences in this area 
is similar, but differs between the two 
amplified regions (16 X on the X com- 
pared to 60 X on the third chromo- 

In the flanking regions the extent of 
amplification decreases roughly by a 
factor of two every 10-15 kb. At the 
level of resolution of these studies the 
decrease appears to be both continuous 
and more or less uniform. However, 
amplification may fall off somewhat 
more rapidly on the third chromosome 
than on the X chromosome. The ampli- 
fied domain surrounding the s36-l and 
s38-l genes is contained within an ap- 
proximately 90-kb cloned region. Not 
all the amplified sequences from the 
66D region have yet been isolated, 
since sequences from one end still un- 
dergo about threefold amplification. 
By extrapolation, however, a total re- 
gion of only about 100 kb is estimated 
to undergo differential replication. 

One model consistent with these ob- 
servations envisions multiple rounds 
of initiation at a specific site or sites in 
the central region of a domain. If elong- 
ation occurs bidirectionally and does 
not terminate or pause at a specific 
site, then changes in the restriction 
fragments generated by the amplified 
sequences would be difficult to detect, 
because fragments containing forks 
would migrate heterogeneously on 
gels. A model of this type has been pro- 
posed by Laird (Cell, 22, 869-874, 1981) 
on the basis of theoretical and struc- 
tural studies of polytene chromosomes. 
Multiple fork structures are present in 
replicating DNA from some bacterio- 
phages, but have never been observed 
in extensive studies of Drosophila em- 
bryo DNA replication. We are cur- 

rently attempting to enrich follicle-cell 
DNA for replicating amplified DNA 
and to examine these preparations in 
the electron microscope. Direct obser- 
vation of amplified DNA may allow us 
to differentiate between this model and 
other possibilities that cannot be com- 
pletely eliminated on the basis of previ- 
ous experiments. 

Genetic Analysis of Amplification 
and Expression 

A. Spradling and B. Wakimoto 

Mutations that disrupt specific as- 
pects of chorion gene amplification and 
expression can provide insight into 
regulatory mechanisms. In last year's 
Report, we discussed the ocelliless (oc) 
mutation (an inversion with break- 
points at 7F1-2 and 8A1-2), and several 
other mutations which alter the normal 
functioning of genes in the 7F1-2 re- 
gion. Of particular interest are those 
associated with chromosome rearrange- 
ments such as In(l)oc, since these 
sometimes show "position effects" 
which we suspect may result from 
disturbances in DNA replication. 

The analysis of the effect of this 
inversion on chorion gene amplification 
began last year and is now largely com- 
plete. We cloned the Eco RI fragment 
containing the distal breakpoint from a 
library of genomic DNA prepared from 
ocelliless ovary DNA. We used this 
fragment to select wild-type genomic 
clones derived from the region of the 
proximal inversion breakpoint. It was 
then possible to carry out a study of 
the effect of the inversion on both prox- 
imal and distal sequences (see Fig. 53). 

These studies have confirmed and ex- 
tended our previous conclusions (Year 
Book 79). Following inversion, the nor- 
mal domain of amplification (Fig. 51) is 
split into two parts. Only the portion 
that is rearranged near 8A continues to 
amplify in follicle cells. Presumably the 
distal part of the amplified region has 
lost sequences which are required in 



10 kb 

a b 

d e f g 



In (I) oc 

i j k Imnopqrst u 

1 I I I II 1 1 i m i — i — I I I II I I II I I 

2203 15 I EcoRI 

ttt n n — i — mi 





1813 704 104 




2203 1406 15 


e *% 

•m» t •• dm mm 


mo mo mo eamo 

mo mo 

I6 r 

Fig. 53. The effect of In(l) ocelliless on X-chromosome gene amplification. A map of the Eco RI 
sites in the X-chromosome chorion gene cluster (see Fig. 51) and in the region of the proximal break- 
point of the ocelliless inversion is shown. Eco RI fragments greater than about 1.5 kb are indicated 
by lower case letters, as shown. Below, the sequences contained in the genomic inserts of the wild- 
type Drosophila clones used as probes to assay amplification are indicated with horizontal lines. 
DNA from each clone was labeled with 32 P and hydribized to Eco Rl-digested DNA from ocelliless 
adult males (m) or from the ovaries of adult homozygous ocelliless females (o) which had been elec- 
trophoresed on a 0.8% agarose gel and transferred to a nitrocellulose filter. In the case of \0101, 
hybridization was also to equal amounts of DNA from wild-type embryos (e) and stage-13 egg 
chambers (a). The ratio of male DNA to ovary DNA was V2 (1406), 1 (1813, 704, 104), 2 (15), or 5 
(0101). The extent of amplification was calculated by determining the intensity ratio of each frag- 
ment between amplified (ovary) DNA and unamplified (male) DNA and then correcting for the dif- 
ferences in total DNA present calculated from Dme8D10 hybridization (z). (In the case of X2203, nor- 
malization was based on the absolute intensity of hybridization, since male DNA was not hybridized 



the cis configuration for amplification. 
The gene dosage of this region, which 
includes several follicle cell RNA tran- 
scripts, is therefore reduced about 
16-fold and may account for the defects 
in eggshell structure that lead to the 
sterility of homozygous ocelliless fe- 
males. Amplification of the inverted 
s38-l and s36-l genes still occurs, but 
the replication of this chromosomal re- 
gion is reduced. In the inverted orienta- 
tion, amplification continues to involve 
a chromosomal domain centered in the 
region of the coding sequences. As a re- 
sult of the inversion, however, one of 
the flanking regions is composed of se- 
quences from the 8A region that are 
normally not amplified in follicle cells. 
In ocelliless homozygotes, amplifica- 
tion "spreads" into these sequences for 
a distance of at least 15 kb. Thus a 
single inversion increases or decreases 
the gene dosage of various chromo- 
somal sequences surrounding its break 
points. The simplest interpretation is 
that amplification requires a specific 
sequence in cis, possibly for recogni- 
tion as an initiation site. Replication 
from this site proceeds bidirectionally, 
giving rise to the observed domains. 
The maximum distance over which a 
breakpoint could exert genetic effects 
would thus depend on the size of the 
replicative domain which was dis- 
rupted. The detailed effects of In(\) 

ocelliless on the temporal regulation of 
gene expression are currently under in- 

Further studies to define the chromo- 
somal sequences that control the time 
and extent of amplification of the 
X-linked genes will use the following 
genetic approaches. First, we will ex- 
tend the analysis of oc by selecting for 
x-ray-induced mutations that revert 
the oc female sterility to fertility. The 
recovery and characterization of these 
revertants may allow us to study 
events which restore amplification to 
near normal levels. It is of interest to 
determine if reversion requires the ex- 
act rein version of the 7F1, 2-8 A region 
or if less precise mechanisms including 
second site mutations are sufficient. 

Our second approach involves using 
conventional genetic means to recover 
a series of new mutations and break- 
points in the chorion gene region. 
Normal X chromosomes will be muta- 
genized with x rays, then tested in het- 
erozygous combination with a deletion 
of the 7E-8A polytene interval for the 
presence of lethal or visible mutations. 
By examining the effects of a series of 
different mutations and breakpoints 
on the structure of the DNA, the cho- 
rion RNAs, and the proteins, we can 
more precisely identify chromosomal 
regions essential for chorion gene 
amplification and expression. 

with this probe.) The extent of amplification of each fragment is plotted directly beneath it. The 
hatched bars indicate the observed amplification in ocelliless DNA. Open bars show the level of 
amplification of the corresponding region in wild-type DNA (Spradling, 1981). "y" is the 1.7-kb wild- 
type Eco RI fragment of \0101 containing the proximal ocelliless breakpoint. 





S. L. McKnight, E. R. Gavis, and R. C. Kingsbury 

During the past three years we have 
conducted studies of the thymidine 
kinase (tk) gene of herpes simplex virus 
(HSV). We first isolated the gene by 
molecular cloning (Year Book 78), then 
characterized its structure by resolv- 
ing the DNA sequence and mapping 
the tk messenger RNA (Year Book 79). 
These structural studies have shown 
the gene to be a small protein-coding 

gene. Its size, as measured from the 
site where transcription starts to the 
site where poly-A is added onto tk 
mRNA, is 1308 nucleotides. This seg- 
ment of HSV DNA is complementary 
to tk mRNA throughout its length, in- 
dicating that the tk gene is not inter- 
rupted by intervening DNA sequences. 
Figure 54 summarizes our characteri- 
zation of the tk gene. 

■ 9 K 2 

♦ Hi 

5' tkmRNA 


* t 

500 bp 



♦1 +100 





Fig. 54. Structural features of the HSV tk gene, (a) Restriction endonuclease recognition sites sur- 
rounding the herpes simplex virus thymidine kinase gene. Darkened region denotes segment of 
DNA complementary to tk mRNA. (b) Nucleotide sequence surrounding the 5' terminus of the HSV 
thymidine kinase gene, the noncoding strand of the DNA sequence is displayed progressing from a 
Pvu II restriction enzyme recognition site to a position 200 nucleotides internal to the mRNA- 
coding component of the gene. The predominant 5' termini of tk mRNA map to adenosine residues 
+ 1 and +3. The AUG triplet at nucleotide +110 is the translation start codon nearest the 5' ter- 
minus of tk mRNA. Probe DNA for quantitative SI mapping of tk mRNA 5' termini is made by 
cleaving with Bgl II, which cuts the coding strand at nucleotide +56, and Eco RI, which cuts at 
nucleotide —75. 



As a complement to the structural 
studies mentioned above, we have ex- 
amined the expression of isolated tk 
DNA. Two different assays have been 
used to test for expression of the tk 
gene. We initially found that a recom- 
binant plasmid containing a copy of 
the tk gene is capable of transforming 
tk - cells to the tk + phenotype (Year 
Book 78). This observation was consis- 
tent with previous experiments, which 
showed that transformation of tk - 
cells with HSV DNA results from the 
stable integration and expression of a 
donor HSV tk gene (Wigler et al, Cell, 
11, 223, 1977). As a second in vivo ex- 
pression test, we injected the tk gene 
into frog oocyte nuclei. As reported 
last year (Year Book 79), microinjected 
oocytes synthesize authentic viral tk 

During the past year, our studies 
have focused on determining which 
DNA sequences direct RNA poly- 
merase entry onto the tk gene. The ap- 
proach we used follows that which 
Brown and colleagues used to identify 
the transcriptional control region of 
Xenopus 5S RNA genes (Brown et al, 
Year Book 78). Briefly, tk DNA is first 
mutated systematically in vitro. Mu- 
tated genes are then characterized by 
DNA sequencing. Finally, specific tk 
mutants are tested, by oocyte injec- 
tion, for the retention of transcrip- 
tional competence. 

Frog Oocytes Transcribe 
tk DNA Accurately 

Since we sought to use frog oocytes 
to study those components of the tk 
gene required for transcription initia- 
tion, it was important to show that 
oocytes express tk mRNA accurately 
when injected with a wild-type tk gene. 
To test the accuracy of tk mRNA syn- 
thesis in oocytes, RNA isolated from 
injected samples was subjected to a 
standard transcript mapping proce- 
dure. The transcript mapping proce- 
dure we routinely used involved hybrid- 

izing a radiolabeled single-stranded 
DNA fragment to tk mRNA. The probe 
DNA, a 131-nucleotide segment of tk 
DNA progressing from an Eco RI re- 
striction cut to a Bgl II cut, overlaps 
the 5 ' terminus of the tk mRNA mole- 
cule. When the RNA/DNA hybrid mol- 
ecules that were derived from authentic 
tk mRNA are exposed to the single 
strand specific nuclease, S^ the 131- 
nucleotide DNA probe is trimmed to a 
length of 54-56 nucleotides (Year Book 
79). Figure 55 shows an autoradio- 
graphic exposure of a sequencing gel 
used to size Sj -resistant DNA frag- 
ments derived from the hybridization 
of the 131-nucleotide Eco RI/Bgl II- 
probe DNA fragment to various RNA 
samples. As the gel exposure shows, tk 
mRNA synthesized in oocytes is iden- 
tical at its 5 ' terminus both to authen- 
tic tk mRNA (synthesized in virally in- 
fected mammalian cells) and to tk 
mRNA derived from a transformant 
cell line that carries a single copy of the 
HSV tk gene. 

Construction of Deletion 
Mutants of the tk Gene 

Our initial method of identifying nu- 
cleotide sequences necessary for accur- 
ate expression of the tk gene involved 
preparing and assaying two systematic 
sets of deletion mutants. Deletion mu- 
tagenesis was accomplished by an en- 
zymatic method using exonuclease III 
and S x nuclease (Brown et al, Year 
Book 78). A set of mutants that lack 
progressively greater extents of 5' 
flanking DNA in a 5 ' -> 3 ' direction (5 ' 
deletion mutants) was prepared. These 
5 ' mutants were derived by enzymatic 
digestion starting at a Bam HI restric- 
tion site approximately 680 nucleotides 
upstream from the tk structural gene 
(Fig. 54). Using similar methods, a 
systematic set of deletion mutants pro- 
gressing in a 3 ' -* 5 ' direction (3 ' dele- 
tion mutants) was prepared by en- 
zymatic digestion starting at a Bgl II 
restriction site 56 nucleotides within 



Single Copy. Infected Injected 

Transformant Cells Oocytes 

40 20 10 40 20 10 40 20 10 

III .lit lit 

a» « 



122 - «m 
MO - ** 

90 - 

76 - 

67 - 


Fig. 55. SI nuclease maps of tk mRNA synthesized in HSV-infected cells, tk + transformant 
cells, and microinjected frog oocytes. Figure shows an autoradiograph exposure of a 14% 
polycarylamide electrophoresis gel used to size Sl-resistant tk probe DNA. RNA prepared from 
HSV-infected cells, tk+ transformant cells, and frog oocytes injected with tk plasmid DNA was 
hybridized to an excess of radiolabeled, 131-nucleotide Bgl II/Eco RI, single-strand probe DNA. The 
probe DNA overlaps the 5' terminus of the tk gene (see Fig. 54). RNA/DNA hybrids were digested 
with 10, 20, or 40 units/ml SI nuclease. Sl-resistant DNA was sized by electrophoresis on a 14% 
polyacrylamide sequencing gel. Numbers to the left denote the size of molecular weight marker DNA 
fragments in nucleotides. 



the tk structural gene (Fig. 54). The 
precise deletion end point of each mu- 
tant was resolved by DNA sequencing. 
Deletion mutants were identified with 
respect to the direction of deletion (5 ' 
or 3 ') and to the nucleotide at which en- 
zymatic deletion terminated with re- 
spect to the putative start site of tk 
mRNA synthesis (nucleotide +1, as 
shown in Fig. 54b). 

Expression of 5' Deletion 
Mutants in Frog Oocytes 

We injected 10-ng supercoiled DNA 
of each 5' deletion mutant into the 
nuclei of 100 oocytes. Following a 
24-hour incubation interval at 20 °C, we 
isolated and assayed RNA for the pres- 
ence of authentic tk mRNA by S : nu- 
clease mapping. Figure 56 shows an au- 
toradiographic exposure of a sequenc- 
ing gel used to resolve transcript maps 
of RNA synthesized from various 5' 
deletion mutants. As the gel exposure 
of Fig. 56 clearly shows, the amount of 
authentic tk mRNA synthesized by 
5'— 85 is drastically reduced as com- 
pared with that of 5 ' —109 and with all 
mutants that retain > 109 nucleotides 
of contiguous 5 ' flanking DNA. 

The reduction in tk mRNA expres- 
sion observed at the —109/— 85 bound- 
ary can be quantitated by measuring 
with scintillation spectrometry the 
amount of radioactive DNA in excised 
gel bands. As Table 3 shows, the amount 
of authentic tk mRNA synthesized by 
5' deletion mutants that retain >109 
5' flanking nucleotides is roughly 50 
times that of ones that retain between 
32 and 85 contiguous 5 ' flanking nucle- 
otides. These data are supported by 
two independent observations. First, 
the amount of enzymatically active tk 
protein made in oocytes injected with 
5'— 109 is 20-40 times greater than 
that made by 5 ' -85 (Table 3). Second, 
mutants that retain >109 5' flanking 
nucleotides transform tk - mouse cells 
to the tk + phenotype 20 times more ef- 
ficiently than do those that retain < 85 

contiguous 5' flanking nucleotides 
(McKnight, Gavis, Kingsbury, and 
Axel, manuscript submitted). 

These results indicate that a region 
of DNA upstream from the tk struc- 
tural gene has a profound effect on the 
efficiency of transcriptional expres- 
sion. The 5 ' boundary of this region is 
located somewhere between the dele- 
tion end points of 5 ' —109 and 5 ' — 85. 
Since the deletion isolate 5'— 95 re- 
tains a significant fraction of the wild- 
type transcription potential, we ten- 
tatively assign the 5 ' boundary of this 
transcriptional control region to nucle- 
otide -100. 

Expression of 3' Deletion 
Mutants in Frog Oocytes 

The 3 ' deletion mutants constructed 
for this study lack most, if not all, of 
the tk structural gene. The 5 ' flanking 
DNA retained by this series of dele- 
tions extends from a fixed site 480 
nucleotides upstream from the 5' ter- 
minus of the tk structural gene to 
varying locations in and around the pu- 
tative transcription start site. The no- 
menclature we use to denote 3' 
mutants refers to the site at which the 
synthetic oligonucleotide linker mole- 
cule was ligated to the deletion end 
point. For example, 3 ' +7 refers to a 3 ' 
mutant that retains 480 nucleotides of 
5' flanking DNA as well as 7 nucleo- 
tides of tk mRNA-coding DNA. 

Since our 3' deletion mutants lack 
the tk structural gene, we have ana- 
lyzed their ability to promote expres- 
sion of plasmid sequences juxtaposed 
to each deletion mutant end point. This 
analysis was accomplished by a hybrid- 
ization assay using the two single 
strands of a discrete DNA fragment 
derived from the tetracycline gene of 
pBR-322. RNA isolated from oocytes 
that were injected with individual 3' 
deletion mutants was hybridized to 
either the "coding" or "noncoding" 
strand of a radiolabeled 53-nucleotide 
Hae II restriction fragment. The Hae 


>£t<y, II 

</> <r tf O - - - - -- - -- - - 

X Q.-+- o m in ^ m m m i/) m in in id mm 













^d MMS^ 




Fig. 56. SI nuclease transcript maps of tk mRNA synthesized in frog oocytes injected with 5' 
deletion mutants. Figure shows an autoradiograph exposure of a 14% polyacrylamide elec- 
trophoresis gel used to size Sl-resistant tk probe DNA. For each experiment, 5 fig of RNA was 
hybridized with 5-ng radiolabeled Bgl II/Eco RI probe DNA. RNA/DNA hybrid molecules were 
digested with 10 units/ml SI nuclease. Lane labeled HSV shows transcript map of authentic tk 
mRNA. Lane labeled parental shows transcript map of RNA synthesized in oocytes in response to 
microinjection of ptk/A3' —1.13, the parental tk isolate from which 5' deletion mutants were derived. 
Lane labeled -faA shows transcript map of RNA synthesized in oocytes injected with ptk/A3' —1.13 
and 1 uglml a-amanitin. Lane labeled oocyte shows transcript map of RNA prepared from unin- 
jected oocytes. Remaining lanes show transcript maps of RNA made in oocytes in response to 
various 5' deletion mutants. All transcription reactions were assayed in oocytes isolated from a 
single female frog. Numbers to right denote sizes of molecular-weight-marker DNA fragments in 



TABLE 3. Functional Assays of 5' Deletions 
in Oocytes and Fibroblasts 

tk Enzymatic 

Activity (units 


above 5 units/ 


tk mRNA* 

ml Oocyte 


(cpm X 10 _1 ) 





5' -182 



5' -148 



5' -119 



5' -109 



5' -95 



5' -85 



5' -78 



5' -70 



5' -56 



5' -46 



5' -32 



5' -18 


5' -6 


♦Represents Cherenkov counts per minute 
above the background (45 cpm) of a gel slice cor- 
responding to the 52-57 nucleotide range ex- 
cised from a lane loaded with probe DNA that 
had been hybridized to RNA from uninjected 
oocytes prior to Sj nuclease digestion. 

II fragment is located within the tetra- 
cycline gene of pBR-322 some 121 nu- 
cleotides from the junction at which tk 
5' flanking DNA is joined to pBR-322 
DNA. "Coding" strand refers to the 
Hae II DNA strand that replaced the 
tk mRNA-coding DNA strand. Simi- 
larly, "noncoding" refers to the Hae II 
DNA strand that replaced the non- 
coding strand of the tk gene. 

When 3 '+7 plasmid DNA is in- 
jected into oocytes, it expresses at 
least ten times more RNA complemen- 
tary to the Hae II coding strand than 
it does to the noncoding strand. This 
transcriptional asymmetry is not ob- 
served when pBR-322 alone is intro- 
duced into oocytes, nor is it observed 
when 3' +7 is injected in the presence 
of 1 pig/ml a-amanitin. These quanti- 
tations are derived from nucleic acid 
hybridizations using an excess of 
radiolabeled Hae II-probe DNA. 
Following annealing, RNA/DNA 
hybrids are exposed to Sj nuclease, 

precipitated on filter discs with TCA, 
and counted by scintillation spectrom- 

3 ' deletion mutants extending as far 
as 3'— 52 are capable of promoting 
asymmetric transcription of plasmid 
sequences juxtaposed to tk 5 ' flanking 
DNA (Table 4). The isolate 3 ' -69 also 
directs accentuated expression of the 
Hae II coding-strand sequence, yet at a 
level only 25% that of 3' -52. The 3' 
deletions that remove 82 or more 5' 
flanking nucleotides do not promote 
expression of the plasmid DNA strand 
that replaced tk mRNA coding-strand 
DNA (Table 4). 

These experiments indicate that a 
segment of DNA at least 52 nucleo- 

TABLE 4. Hybridization Analysis of Tran- 
scripts Synthesized by 3' Deletion Mutants 


Hae II* 

S j -Resistant 

(cpm X 10 ~ 2 ) 

3' +7 



3' +7 










3' -2 



3' -7 



3' -16 



3' -21 



3' -27 



3' -37 



3' -52 



3' -69 



3' -84 



3' -105 



*Hae II slow-migrating single-strand probe 
was labeled to a specific activity 3.1 times that 
of the fast-migrating strand due to the asym- 
metric distribution of adenosine and cytosine 
residues replaced by Exo Ill/reverse transcript- 
ase labeling. 

f5-/ug oocyte RNA was annealed to 1-ng 
single-strand probe DNA for 2 h at 65 °C. 
Hybrids were diluted 10-fold into S : nuclease 
buffer, digested for 30 min at 20° using 50 
units/ml- S, nuclease, and assayed for nuclease 
resistance Dy TCA precipitation and scintilla- 
tion spectrometry. 

ta-amanitin was co-injected at a concentra- 
tion of 1 /ig/ml. 



tides upstream from the putative 
transcription start site of the tk gene 
can promote transcriptional expression 
by RNA polymerase form II in the 
complete absence of mRNA-coding se- 
quences. In concert with the analyses 
of 5' deletion mutants discussed ear- 
lier, these data delimit the boundaries 
of a transcriptional control region oper- 
ative in vivo. A segment of tk 5 ' flank- 
ing DNA between 109 and 52 nucleo- 
tides upstream from the tk structural 
gene appears to serve a critical role in 
ensuring quantitative expression of se- 
quences that lie 3 ' to the region. 

The quantitative hybridization as- 
says described in the preceding para- 
graphs do not resolve the position at 
which transcription starts on the vari- 
ous 3' deletion mutants. To approach 
this question, we used a primer exten- 
sion assay to map the 5' termini of 
transcripts synthesized in oocytes 
from the various 3' deletion mutants. 
The same 53-nucleotide Hae II coding- 
strand probe used for quantitating 
transcripts promoted by 3' deletion 
mutants was annealed to RNA synthe- 
sized from the different 3' deletions. 
RNA/DNA hybrids were then exposed 
to reverse transcriptase in the presence 
of excess deoxyribonucleoside triphos- 
phates. This procedure extended the 
length of the primer DNA to the 5 ' ter- 
minus of the RNA molecule to which it 
was annealed. 

Figure 57 shows an autoradiographic 
exposure of a sequencing gel used to 
map the 5' termini of transcripts syn- 
thesized in oocytes from the various 3 ' 
deletion mutants. Two aspects of these 
data are noteworthy. First, the same 3 ' 
mutants that promoted asymmetric 
expression of the pBR-322 Hae II frag- 
ment, as assayed by quantitative hy- 
bridization, also synthesized RNA cap- 
able of extending the length of the Hae 
II primer. Second, the largest quantita- 
tively significant extension product 
derived from RNA synthesized by 
seven of the 3 ' deletion mutants (3 ' +7 
through 3'— 37) was inversely propor- 

tional to the extent of 3 ' — 5 ' deletion. 
For example, the largest extension 
product derived from RNA synthe- 
sized by 3' +7 was approximately 185 
nucleotides in length. 3 ' — 2, which was 
deleted in a 3 ' — 5 ' direction nine nucle- 
otides beyond 3 '+7, produced trans- 
cripts that extended the Hae II primer 
to a length of approximately 176 nucle- 
otides (Fig. 57). Much the same pattern 
was observed for RNA synthesized by 
3' -7, 3' -16, 3' -21, 3' -27, and 
3' -37. 

We interpret these data in the follow- 
ing way. When 3' +7 plasmid DNA is 
microinjected into oocytes, it promotes 
synthesis of transcripts that initiate at 
nucleotide +1 of the tk structural 
gene. Elongation of the nascent RNA 
chain progresses through the seven 
nucleotides of tk structural DNA, and 
continues through plasmid sequences 
past the 53-nucleotide Hae II frag- 
ment. Thus, when assayed by primer 
extension, a 186-nucleotide product is 
observed (7 nucleotides tk DNA +126 
nucleotides pBR-322 DNA from Bam 
HI site to Hae II site + 53 nucleotides 
Hae II primer = 186). When 3' -2, 
3'-7, 3'-16, 3'-21, 3'-27, and 
3 ' — 37 are transcribed in oocyte nuclei, 
they all initiate transcription at or very 
near the position formerly occupied by 
nucleotide +1 of the tk structural 
gene. These data suggest that se- 
quences upstream from nucleotide —37 
not only promote transcriptional ex- 
pression of the HSV thymidine kinase 
gene in a quantitative sense, but also 
specify the approximate start site of 


We have shown that the HSV 
thymidine kinase gene is accurately ex- 
pressed in Xenopus laevis oocytes. By 
constructing, characterizing, and assay- 
ing two systematic sets of deletion 
mutants of the tk gene, we have re- 
solved the location of a transcriptional 
control region. The control region 



cm r- 

— r- 


t I 












rOrOrO roro fOrOrOrOfO 




^^" m^v ^n^ ^^ ^^^ 
•''""■"t ^^ ,w ^"* ^^ 








Fig. 57. Primer extension transcript maps of RNA synthesized in frog oocytes injected with 3' 
deletion mutants. Figures shows an autoradiographic exposure of a 10% polyacrylamide elec- 
trophoresis gel used to size reverse transcriptase extension products generated from various RNA 
samples. For each experiment, 5 ng of RNA was hybridized with 1 ng of radiolabeled pBR322 Hae II- 
coding-strand probe. RNA/DNA hybrids were extended using 100 units/ml AMV reverse transcrip- 
tase. Deletion mutants tested are indicated above gel lanes. All transcription reactions were assayed 
in oocytes isolated from a single female frog. Numbers to right denote sizes of molecular weight 



reaches from a position approximately 
100 nucleotides upstream from the tk 
gene to a position approximately 40 
nucleotides upstream from the gene. 
This 60-nucleotide segment of tk 5' 
flanking DNA is capable of promoting 
expression of sequences 3 ' to it in the 
absence of both the tk mRNA-coding 
sequences and the 40 nucleotides 
closest to the putative transcription 
start site. Moreover, expression medi- 
ated from this transcriptional control 
region appears to start at a specified 
location some 70 nucleotides 3' to its 

How Does tk mRNA Avoid the 
Splicing Requirement? 

Hugh Pelham 

One interesting feature of the herpes 
tk gene is that it lacks intervening se- 
quences (McKnight, Year Book 79). 
Most other genes of higher eukaryotes 
contain intervening sequences, and in 
the cases that have been examined it 
seems that a splicing even is essential 
for the production of stable cytoplas- 
mic mRNA. It could be that the tk 
transcript contains some special fea- 
ture allowing it to bypass this splicing 
requirement. If so, other sequences in- 
troduced into the tk mRNA might also 
be stabilized, and by deletion analysis 
it should be possible to identify the tk 
sequences involved. 

Hybrid genes were constructed con- 
sisting of rabbit /3-globin cDNA in- 
serted into the 5' untranslated region 
of the tk mRNA sequence. As a con- 
trol, the globin cDNA was joined di- 
rectly to the tk promoter, using one of 
S. McKnight's deletion mutants that 
lacked all the tk mRNA sequences. The 
tk gene and the hybrid genes were in- 
serted into a special pBR322-derived 
vector obtained from M. Botchan. This 
vector, which contained a small portion 
of the SV40 genome, including the 
origin of replication, replicated ef- 
ficiently when introduced into cells 
containing SV40 T antigen. (pBR322 

sequences between bases 1120 and 
2490, which apparently do not repli- 
cate well in animal cells, were deleted 
from the vector.) Cos-7, a monkey cell 
line containing an integrated defective 
SV40 genome which produces T anti- 
gen constitutively, was used as a host- 
cell line. 

The plasmids were introduced into 
the Cos cells either by the calcium 
phosphate co-precipitation method or 
in the presence of DEAE-dextran. Two 
days later, RNA was prepared from the 
cells and tk-derived transcripts were 
detected by Sj mapping. Unfortu- 
nately, the tk promoter appeared to be 
inactive in this assay; the only detect- 
able transcripts were initiated well up- 
stream of the expected position. This 
inactivity may not be artif actual. Ex- 
pression of the viral tk gene in herpes- 
infected cells shows an absolute re- 
quirement for the product of another 
viral gene (Post et al, Cell, 24, 555, 
1981). Nevertheless, the tk gene is ex- 
pressed in cell lines converted from tk~ 
to tk + with the isolated gene; S. Mc- 
Knight has shown that the tk promoter 
is used in such cells. However, these 
cell lines are selected for tk expression, 
and the site of integration of the gene 
may be crucial for its transcription. 

Steven McKnight found that all hy- 
brid genes injected into frog oocytes 
produced abundant transcripts ini- 
tiated at the tk promoter. The oocyte 
environment therefore appears to be 
permissive for gene expression. 

These results suggest two directions 
for future work. Requirements for 
mRNA export from the nucleus may be 
studied using the oocyte-injection 
assay. Although all transcripts derived 
from the tk promoter seem to be stable 
in this system, improper ones may well 
be retained in the nucleus and not 
translated into protein, a prediction 
that should be easy to test. The Cos 
cell system, on the other hand, may 
provide a tool for exploring the require- 
ments for tk-gene activity under more 
restrictive conditions. 



J. G. Gall 

The expression of histone genes dur- 
ing the lampbrush chromosome stage 
of oogenesis in the newt Notophthal- 
mus viridescens was studied by in situ 
hybridization and by SI nuclease map- 
ping of RNA transcripts. Previous 
studies showed that histone gene clus- 
ters are separated by long tracts of a 
tandemly repeated 222-bp sequence 
(satellite 1) and that both gene and 
satellite sequences are transcribed on 
the same lampbrush chromosome loops 
(Diaz et al., Cell, 24, 649-659, 1981; 
Stephenson et al, Cell, 24, 639-647, 
1981; Gall etal, Chromosoma, in press). 
We postulated that transcription be- 
gins at a histone gene promoter, fails 
to terminate at the end of the gene, and 
continues through the rest of the gene 
cluster into the satellite sequences. 
Since the cytoplasmic mRNAs are only 
440-760 nucleotides in length, exten- 
sive processing of the primary tran- 
scripts must occur. Additional evidence 
for this model was obtained by showing 
that a probe consisting entirely of 

sequences between the H4 and H2A 
genes hybridizes strongly to the appro- 
priate lampbrush chromosome loops. 
SI nuclease mapping is being used to 
characterize the histone mRNAs and 
primary transcripts. The 5' ends of the 
cytoplasmic HI and H4 mRNAs have 
been defined by hybridizing appropri- 
ate restriction fragments of cloned 
genomic DNA to RNA from mature oo- 
cytes. The H4 mRNA has a 5' un- 
translated segment approximately 33 
bp in length while that of the HI 
mRNA is about 50 bp. Oocytes store 
such large amounts of histone mRNA 
that SI mapping can be done with 
RNA from single oocytes. Similar ex- 
periments are under way to define the 
primary transcripts in the germinal 
vesicle. If readthrough transcription 
occurs as we believe, only a fraction of 
the nuclear transcripts will have 5' and 
3' ends corresponding to those of the 
mature mRNAs. Most nuclear tran- 
scripts should give full protection of 
probes in SI mapping experiments. 


D. D. Brown, D. F. Bogenhagen, I. J. Jackson, E. Jordan, H. Pelham, S. Sakonju, 
and W. M. Wormington 

We study the 5S ribosomal RNA 
genes with the hope of reconstructing 
their transcriptional control in vitro. 
In previous Year Books we have de- 
scribed the isolation and detailed char- 
acterization of these very simple genes. 
An in vitro assay system has enabled 
us to delimit DNA sequences within 
and near the gene that influence accu- 
rate initiation and termination of tran- 
scription. Initiation is controlled by a 
region in the center of the gene that 
functions by binding a specific tran- 
scription factor. This factor, in turn, 

directs RNA polymerase III to start 
transcription at the correct site. A sim- 
ple sequence sufficient to terminate 
transcription was reported in Year 
Book 79 by D. Bogenhagen. 

With this preliminary information in 
hand, we have begun to study the de- 
velopmental control of 5S RNA genes. 
In this report, D. Bogenhagen shows 
that the transcription complex is ex- 
ceedingly stable. M. Wormington con- 
firms this conclusion with his studies 
of various 5S RNAs transcribed from 
chromatin in cell extracts. The chroma- 



tin is not influenced by the source of 
the extract but continues to support 
transcription of somatic 5S RNA. 
When the chromatin is washed with 
saline, the normally silent oocyte-spe- 
cific genes are transcribed. This gives 
us a source of repressed chromatin 
under the same developmental control 
as it is in the living cell. Therefore, we 
can show that the chromatin contains 
information needed to maintain oocyte 
genes silent, and somatic genes active, 
in somatic cells. The rearrangement or 
removal of some molecules bound to 
the chromatin DNA is sufficient to 
make the repressed oocyte genes acces- 
sible to the extract, which can then 
transcribe the previously silent genes. 
The transcription complex formed by 
the addition of genes to an extract in 
vitro resembles this chromatin; it can 
support many rounds of RNA synthe- 
sis without dissociating. This stability 
is expected for eukaryotic transcrip- 
tion where genes must remain active 
(or inactive) for long periods of time. 

D. Bogenhagen and M. Wormington 
have found a quantitative difference in 
the transcription of oocyte and somatic 
genes. Two or three base differences 
in the control region are enough to 
weaken an oocyte gene's ability to bind 
the transcription factor. However, this 
difference in binding constant is not 
sufficient to account for the genes' de- 
velopmental control. 

The transcription factor is an extra- 
ordinary protein that binds not only 
the gene but also the gene product (5S 
RNA). Synthesis of the protein occurs 
at a high rate in oocytes. I. Jackson is 
investigating this important part of 5S 
gene control by cloning the gene for the 
factor. H. Pelham has shown that a 
protein related to the oocyte factor, but 
which appears to be larger by its mi- 
gration in gels, is probably the factor 
that functions in somatic cells. No test 
of physiological function to date has 
distinguished these two proteins. 

Finally, a reinvestigation of the nor- 
mal developmental pathway of 5S RNA 

synthesis has revealed the following 
series of events. A massive synthesis 
of oocyte 5S RNA occurs during oo- 
genesis. All RNA synthesis ceases during 
meiosis (in the unfertilized egg and dur- 
ing cleavage), and begins again at 
about the 4000- to 8000-cell stage, just 
before the onset of gastrulation. At 
this stage, the 5S RNA synthesized is 
an equal mixture of somatic and oocyte 
5S RNAs. After another cell division, 
the embryo synthesizes only somatic 
5S RNA. 

Binding of a Transcription Factor 

to the Control Region of the 5S 

RNA Gene 

S. Sakonju 

We continue to study the binding of 
a 40,000-dalton protein that is required 
for transcription of the 5S RNA gene. 
The protein protects the middle of the 
gene, between residues +45 and +97, 
from DNase I digestion. In Year Book 
79, we reported experiments in which 
we examined the binding of this tran- 
scription factor to various deletion 
mutants of 5S DNA. We found that 
the factor would not bind to the control 
region when the gene was deleted from 
the 3' end to residue +80. It would 
bind, however, albeit weakly, to a con- 
trol region that had been deleted from 
the 5' end to as far as residue +74. We 
concluded from these observations that 
the sequence toward the 3' side of the 
binding region is the dominant recogni- 
tion sequence without which binding to 
the 5' region cannot occur. 

In this report, we describe experi- 
ments in which we determined some of 
the residues that interact directly with 
the protein. DNA fragments from 
pXbs 201, containing the somatic-type 
gene and about 50 bp of the flanking 
sequences on both its 5' and 3' ends, 
were end-labeled with 32 P on either the 
coding or noncoding strand. The frag- 
ments were then treated for a short 
time with dimethylsulfate (DMS) so 



that about one residue per fragment, 
on average, was methylated. The N-7 
position of guanine and, to a lesser ex- 
tent, the N-3 position of adenine were 
methylated by DMS. The modified 
DNA fragments were used in a binding 
reaction. Those fragments that, com- 
plexed with the factor were purified by 
immunoprecipitation with antibody 
against the factor and Staphylococcus 
aureus cells. We found that fragments 
methylated at critical contact points 
cannot bind the factor and remain in 
the supernate. Both the pelleted and 
supernatant fractions were cleaved at 
methylated residues in separate re- 
actions. Electrophoresis of the cleaved 
fragments on a polyacrylamide gel con- 
taining 7M urea is shown in Fig. 58. 
Lanes a and d show the total methyla- 
tion patterns of the coding and non- 
coding strands, respectively, of the 
DMS-treated DNA; the extent of 
methylation of a residue is reflected by 
the darkness of the band. Comparing 
the fragments in the "bound" lanes, b 
and e, with the "total" lanes, a and d, 
reveals several bands that are absent 
in the bound lanes. Those same bands 
are enriched in the unbound lanes, c 
and f. The results indicate that the 
methylation of some guanine residues 
interferes with the binding of the fac- 
tor. Those that are absent in the bound 
lane are guanine residues at +70, +71, 
+81, +82, +85, +86, and +89 on the 
noncoding strand. The bands corre- 
sponding to residues +87 on the non- 
coding strand and +91 and +92 on the 
coding strand are present but reduced 
when compared to the control lane, in- 
dicating some interference by methyla- 
tion. We emphasize that the fragments 
were methylated lightly so that, on 
average, only one residue per fragment 
was modified. Therefore, the absence of 
a band on the gel indicates that meth- 
ylation of that residue alone was 
enough to terminate binding by the 
factor. Hence the results confirm the 
conclusion that binding of the protein 
is critically dependent on the sequence 

+ 120 - 

+ 100- 

+ 80 - 


+ 20 


- +60 

+ 60- 



+ 40 - 


' :-'• ■■■: 

Fig. 58. Identification of Xbs 5S gene 
residues which, when in vitro and methylated, 
interfere with the binding of the transcription 
factor. Xbs 201 DNA fragments, labeled either 
on the coding strand (lanes a, b, and c) or the 
noncoding strand (lanes d, e, and f), were cleaved 
at the methylated residues and run on an 8% 
polyacrylamide 7M urea gel. Lanes a and c show 
the total methylation pattern. The methylated 
residues of the fragments that bound to the 
transcription factor are shown in lanes b and e, 
and those that did not, in lanes c and f. The 
numbers in the margins indicate the positions of 
the 5S gene residues. 



at the 3' side of the binding region. The 
methylation experiment also indicates 
that the protein must interact with the 
DNA in the major groove of the helix 
since that is where alkylated deriva- 
tives on the N-7 position of guanine are 

We have also examined the inter- 
action of the protein with the phos- 
phate backbone of DNA. Phosphate 
residues were partially ethylated using 
ethylnitrosourea, and the fragments 
that bound to the protein were sepa- 
rated from those that did not using the 
same method described above for the 
methylation experiment. We found 
that ethylation of the phosphates ad- 
jacent to the same guanine residues 
whose methylation had been found to 
abolish binding also interfered with the 

We did not detect specific interaction 
of the protein with the residues toward 
the 5' side of the binding region in 
these experiments. However, M. Worm- 
ington has shown that, in a transcrip- 
tion system, the somatic-type gene is a 
better competitor for the transcription 
factor than the oocyte-type gene. Fur- 
thermore, by constructing a hybrid 
gene, D. Bogenhagen has shown that 
the differences in competition strengths 
are due to the base differences at resi- 
dues + 53, +55, +56, and/or +62. 
These results strongly suggest that the 
residues located toward the 5' side of 
the transcription-factor binding region 
interact specifically with the protein. 
To show directly that the binding 
strength of the protein is different for 
the two types of 5S genes, we devel- 
oped a competition assay that uses the 
"footprint" technique described in 
Year Book 79. Xbs DNA fragments 
labeled at one end were incubated in a 
binding reaction with an increasing 
amount of cold competitor DNA. The 
DNAs were partially digested with 
DNase I and the fragments were elec- 
trophoresed on a polyacrylamide/urea 
gel. The binding strength of each com- 
peting DNA was quantified by the ex- 

tent to which it inhibited the footprint 
of a labeled fragment of 5S DNA. The 
results are shown in Fig. 59. This assay 
indicates that Xbs DNA is a better 
competitor than Xlt DNA. Thus base 
differences between the somatic- and 
oocyte-type genes located in the 5' part 
of the control region influence the 
strength of their interaction with the 

Construction and Quantitative 

Transciption of Somatic:Oocyte 

Hybrid 5S RNA Genes 

D. Bogenhagen and W. M. Wormington 

Understanding the differential tran- 
scriptional control of the somatic and 
oocyte 5S RNA genes has been a major 
goal of this laboratory for several years. 
Both oocyte and somatic 5S RNA 
genes are transcribed actively in the 
oocyte nuclear extract. However, 

0.2 0.4 0.6 0.8 

Competitor DNA (^.g ) 


Fig. 59. Relative competition strengths of 
somatic-type and oocyte-type 5S DNAs for the 
purified transcription factor. In a binding reac- 
tion, approximately 0.04 pmoles of radioactively 
labeled Xbs 201 DNA fragments were incubated 
simultaneously with the indicated amounts of 
nonlabeled pXbs (open triangles), pXlt (open 
circles), or pBR322 (closed circles) DNA. The 
figure shows the quantitation of a prominent 
DNase I -cleaved band in the binding region of 
the coding strand (see Year Book 79). 



Wormington has shown that the 
oocyte-specific 5S RNA genes compete 
only one-fourth as well as somatic- 
specific 5S RNA genes in the quantita- 
tive transcription assay ( Year Book 79). 
In vitro mutagenesis of a somatic 5S 
RNA gene has revealed that two re- 
gions—the center and the 5' flanking 
sequence— contribute to the gene's 
ability to compete in this assay. The 
somatic and oocyte 5S DNA sequences 
differ significantly within their 5' flank- 
ing regions as well as at several resi- 
dues within their coding sequences. In 
order to determine whether the com- 
petitive weakness of oocyte-type 5S 
RNA genes is due to sequence changes 
in one or the other of these regions, we 
have constructed somatic:oocyte hy- 
brid genes. One of these hybrids con- 
tains oocyte-specific base-pair changes 
in the center of the somatic 5S RNA 
gene, while the other contains somatic- 
specific base-pair changes in the center 
of an oocyte 5S RNA gene. 

In constructing these genes, we used 
the cloning vector Ml 3 mp5, which pro- 
vides a selective marker for cloning in 
the Ml 3 bacteriophage DNA. With 
this vector, recombinant DNA can be 
obtained in large amounts in pure 
single-stranded form from viral parti- 
cles as well as in the duplex replicative 
form. To construct the first hybrid, a 
deletion mutant of the Xbs 5S RNA 
gene was transferred to the M13 vector 
and single-stranded DNA was isolated. 
This single-stranded DNA was used as 
a template for in vitro replication of 
the DNA by DNA polymerase I using 
a restriction fragment from the center 
of the Xlt oocyte 5S RNA gene as 
primer. DNA ligase was used to co- 
valently seal the replicated molecules. 
The products contained several mis- 
paired base residues in the center of the 
gene at sites where the oocyte and so- 
matic sequences differed. Following 
transfection of E. coli, these heterodu- 
plex molecules replicated to yield two 
types of daughter molecules, a pure 
somatic-type daughter containing the 

former template strand, and a somatic: 
oocyte hybrid daughter containing the 
in vitro synthesized strand. This ap- 
proach was patterned after the work of 
Hutchison et at. {J. Biol. Chem., 253, 
6551, 1978), who showed that short 
synthetic DNA fragments could be 
used as primers for repair mutagenesis 
of this type using viral DNA as the 
template. However, the mutation in- 
troduced in their experiment was phe- 
notypically selectable in E. coli. This is 
obviously impossible for cloned 5S 
RNA genes. As an alternative, we took 
advantage of the fact that the somatic: 
oocyte hybrid contains an oocyte-spe- 
cific Hpa II cleavage site at gene resi- 
due 62. Following identification, the 
hybrid construction was confirmed by 
direct sequencing. We also attempted 
to construct a similar mutant contain- 
ing the central sequences of the 
somatic 5S RNA embedded in the oo- 
cyte 5S RNA gene. The first of these 
experiments was unsuccessful, pre- 
sumably because of contaminating 5' 
to 3' exonuclease activity in the com- 
mercial DNA polymerase (Klenow 
fragment). We therefore adopted 
another strategy to construct this hy- 
brid, one that required only a short in 
vitro extension of the primer followed 
by restriction endonuclease cleavage to 
prepare a hybrid fragment. This re- 
striction fragment was then religated 
to the remainder of the Xlt 5S RNA 
gene and recloned in pBR322. 

The sequences of the two hybrid 
genes and the results of competitive 
transcription are shown in Fig. 60. The 
Xbs and Xlt "parental" 5S DNA se- 
quences differ at five residues within 
the intragenic control region. However, 
comparing the Xlt 5S DNA sequence 
to Xlo and Xbo 5S DNA sequences 
shows that only three of these residues 
evidence oocyte-specific base-pair 
changes, while the others are unique to 
the Xlt 5S RNA gene. The transcrip- 
tion results clearly indicate that incor- 
poration of the oocyte-specific base- 
pair changes into the somatic 5S RNA 



5S RNA Gene 


Hybrid 1 
Hybrid 2 


40 50 60 70 80 


G— IX c 

-G— TC- 


-G— TA- 






Fig. 60. Comparison of nucleotide sequences of parental and hybrid 5S RNA genes. Nucleotide 
sequences were determined by chemical sequencing. The sequence from residues 36 to 85 is shown 
for the Xbsl 5S RNA gene; only residues that differ from the Xbs 5S RNA sequence are shown for 
the other 5S RNA genes. The transition from T to C at residue 40 of hybrid 2 represents a novel 
mutation acquired during in vitro manipulations or during propagation in E. coli. Competition 
strength was determined as described by Wormington et al. (Cell, 24, 809-817, 1981) and in Year 
Book 79. 

gene reduces the competition strength 
and vice versa. The distribution of base 
changes implicates the two transver- 
sions at residues 53 and 55 as the cause 
of the weak competition strength of 
the oocyte-type 5S RNA genes. The 
oocyte-specific base-pair change at 
residue 79 and the extensive sequence 
changes in the 5' flanking region ap- 
pear to contribute little to the weak- 
ened competitive strength of the oocyte 
genes. In additional experiments, S. 
Sakonju has shown that the Xlt 5S 
RNA gene also binds the 5S RNA gene 
transcription factor more weakly than 
the Xbs 5S RNA gene. These results 
provide a further correlation between 
genetic and physical studies to illus- 
trate the importance of the specific 
transcription factor in the expression 
of 5S RNA genes. 

A Stable 5S DNA Transcription 
Complex Can Be Assembled in vitro 

D. Bogenhagen and E. Jordan 

In the last three years we have 
described experiments using recombi- 
nant DNA technology to delimit the 
nucleotide sequences required for 
faithful initiation and termination of 
5S RNA transcription in vitro. These 

studies have involved the construction 
of recombinant 5S RNA genes that en- 
code transcripts distinguishable from 
wild-type 5S RNA by gel electro- 
phoresis. This allows mixtures of two 
templates to be transcribed together in 
one extract so that relative transcrip- 
tion activities can be determined. The 
first such experiments, reported by 
Wormington and Jordan (Year Book 
79), showed that one mutant 5S RNA 
gene competes with another for tran- 
scription factors. This competition 
largely reflects binding of the positive 
transcription factor described by Roed- 
er's laboratory (Engelke et al, Cell, 19, 
717-728, 1980) and by Pelham {Year 
Book 79) to the intragenic control re- 
gion, although the 5' flanking sequence 
also contributes to the competition 

We have subsequently observed that 
the ability of one gene to compete for 
the transcription of another is greatly 
increased if it is preincubated in the 
oocyte nuclear extract prior to the ad- 
dition of the second gene. The pro- 
nounced effect of preincubation on 
transcription competition between a 
wild-type 5S RNA gene and a maxi- 
gene 5S DNA mutant is illustrated in 
Fig. 61. When the two plasmids were 
mixed before addition to the oocyte 



12 3 4 5 

nuclear extract, they transcribed with 
equal efficiency, since the sequence 
modification of the maxigene did not 
affect sequences involved in transcrip- 
tion initiation. However, when the wild- 
type gene was added to the extract 15 
minutes before the maxigene, virtually 
all of the transcripts produced were 
wild-type 5S RNA. We routinely ob- 
served preferential transcription of the 
first template added to the extract. If a 
sufficient concentration of the first 
template was added, transcription of 
template added later could be com- 
pletely inhibited, as seen in lane 4 of 
Fig. 61. Thus, transcription factors 
bind exclusively to the first DNA tem- 
plate added to form a transcription 
complex that is stable for many rounds 
of transcription without exchanging to 
a second template. Table 5 shows that 
the preferential transcription of the 

TABLE 5. 5S DNA Transcription Complexes Are 
Stable for Many Rounds of Transcription* 

Transcription Efficiencyf 

Fig. 61. Formation of exclusive 5S DNA 
transcription complexes. An autoradiogram of 
a polyacrylamide gel of in vitro-synthesized 
32 P-labeled RNA is shown. The arrow denotes 
the position of wild-type 5S RNA. Each reaction 
contained 10-//g/ml maxigene 5S DNA. Lane (1), 
maxigene 5S DNA alone; lane (2), maxigene 5S 
DNA added 15 min after addition of 10-/*g/ml 
wild-type 5S DNA; lane (3), same as 2, but with 
simultaneous addition of templates; lane (4), 
maxigene 5S DNA added 15 min after addition 
of 20-/xg/ml wild-type 5S DNA; lane (5), same as 
lane 4 but with simultaneous addition of tem- 


1-2 3-4 
hours hours 


Reaction A. Comixture 
wild-type Xbs 5S DNA 3.0 3.5 
maxigene Xbs 5S DNA 3.0 2.3 

Reaction B. Sequentially added 
wild-type Xbs 5S DNA 6.9 7.9 
maxigene Xbs 5S DNA 0.05 0.14 



*Each reaction contained equal amounts (15 
jig/ml each) of two plasmids bearing wild-type 
Xbs 5S DNA or maxigene Xbs 5S DNA. In reac- 
tion A, the two templates were added simul- 
taneously. In reaction B, wild-type Xbs 5S DNA 
was added 15 min before maxigene Xbs 5S 
DNA. After 15 min continued incubation in the 
nucleotide-depleted oocyte extract, transcrip- 
tion was initiated by addition of nonradioactive 
nucleoside triphosphates. Aliquots were re- 
moved after 1, 3, or 5 hours for pulse labeling 
with a 32 P-GTP. The two transcripts were re- 
solved by gel electrophoresis. The incorporation 
of label was quantitated by counting the excised 
RNA bands and used to calculate transcription 

fDuring 1-hour pulse labeling (transcripts/ 



first template persisted for at least six 
hours in vitro. During this interval, the 
first template was transcribed approxi- 
mately 40 times, while an equal amount 
of the second template was not tran- 

In previous transcription experi- 
ments, we observed that a lag period of 
approximately 15 minutes was re- 
quired before 5S DNA added to the 
oocyte nuclear extract was efficiently 
transcribed. This lag period corre- 
sponds to the time required to assem- 
ble the stable transcription complexes. 
The multicomponent assembly reaction 
required to construct stable transcrip- 
tion complexes also explains the find- 
ing that transcription reactions are ex- 
tremely sensitive to extract dilution 
prior to addition of DNA. In contrast, 
after assembly in a concentrated ex- 
tract, 5S DNA transcription complexes 
are quite resistant to dilution. When 
the extract was diluted 16-fold, the 
rate of 5S RNA synthesis decreased 
only 30% after assembly of the com- 

The transcription of 5S RNA genes 
by RNA polymerase III requires a 
number of additional transcription fac- 
tors. One of these has been character- 
ized previously (see Year Book 79). 
This protein, a positive transcription 
factor, binds to the intragenic control 
region of 5S RNA genes. We have 
sought to determine the role of this 
protein in formation of 5S DNA tran- 
scription complexes. We treated the 
oocyte nuclear extract with antibodies 
to the transcription factor coupled to 
Sepharose beads in order to remove 
most of the factor. The factor-depleted 
extract synthesizes very little 5S 
RNA. By varying the timing both of 
readdition of the factor and of ordered 
addition of two 5S DNA templates, we 
showed that the factor is required at an 
early step in assembly of the transcrip- 
tion complex. Presumably other tran- 
scription factors and polymerase III 
do not bind stably to naked 5S DNA, 
but rather recognize the factor:5S DNA 

complex. The observation that tran- 
scription complexes are stable does not 
rule out the possibility that some com- 
ponents of the transcription complex 
may dissociate from the DNA between 
rounds of transcription. However, if 
any factors or the RNA polymerase 
itself are free to dissociate between 
rounds of transcription, they must re- 
bind with greater avidity to the resid- 
ual DNA-protein complex than to 
another naked competitor DNA. 

The remarkable stability of 5S DNA 
transcription complexes assembled in 
vitro is equaled in vivo. Given the 
complexity of the Xenopus genome, 
the formation of stable transcription 
complexes would appear to be more ef- 
ficient than de novo reassembly of the 
complex for each individual transcrip- 
tion event. These results have altered 
our view of transcriptional regulation 
of the somatic and oocyte 5S DNA 
families during development. The 
assembly of one family of 5S RNA 
genes into active transcription com- 
plexes would obviate the need to dis- 
criminate between somatic and oocyte 
5S RNA genes at each round of tran- 

Chromatin Transcription Retains 

Developmental Control of Xenopus 

5S RNA Genes in vitro 

M. Wormington 

In Year Book 79 we described how 
the in vitro transcription of Xenopus 
5S RNA genes and of deletion mutants 
of these genes have been quantitated 
by assays that measure the efficiency 
of transcription and the ability to com- 
pete with transcription of a wild-type 
5S RNA gene. These assays revealed 
that the major-oocyte and trace-oocyte 
5S RNA genes from X. laevis are tran- 
scribed as efficiently as somatic 5S 
DNAs, but that they compete for tran- 
scription factors only one-fourth as 
well in an oocyte nuclear extract (Worm- 
ington et al, Cell, 24, 809, 1981). We 



have further exploited this competition 
assay to identify the sequence differ- 
ences within the intragenic control re- 
gion responsible for reduced oocyte 
competition strength. 

We also used extracts prepared from 
Xenopus tissue culture cells to com- 
pare the relative template activities of 
oocyte and somatic 5S DNAs. Tran- 
scription of cloned 5S RNA genes in 
somatic extracts may be dependent 
upon a protein that is similar to the 
previously characterized oocyte tran- 
scription factor (Year Book 79) but 
that has a larger molecular weight (Pel- 
ham et al, Proc. Nat. Acad. Sci. USA, 
78, 1760, 1981). Competition assays 
between oocyte and somatic 5S DNAs 
indicate that somatic 5S RNA genes 
compete four times better than oocyte 
5S DNAs in somatic cell extracts, just 
as they do in oocyte nuclear extracts. 
Thus, the presence of two distinct 5S 
gene transcription factors in addition 
to the fourfold difference in competi- 
tion strength cannot entirely account 
for the developmental inactivation of 
oocyte-type 5S RNA genes in somatic 

Therefore, we have initiated studies 
on the transcription of endogenous 5S 
RNA genes in the form of chromatin 
templates isolated from oocytes and 
various somatic cells. Purification of 
5S RNA either from Xenopus tissue 
culture cells pulsed briefly with 32 P0 4 
or from isolated nuclei incubated with 
32 P-ribonucleoside triphosphates, and 
subsequent fingerprinting, reveals that 
this 5S RNA is entirely somatic. 
Chromatin prepared from hypotoni- 
cally disrupted tissue culture or eryth- 
rocyte nuclei is used as a template for 
in vitro transcription in oocyte- and 
somatic-derived extracts. Upon incu- 
bation in either extract, somatic chro- 
matin directs the synthesis of only so- 
matic-type 5S RNA, indicating that 
oocyte-type 5S RNA genes are main- 
tained in their developmentally inacti- 
vated state. This control is retained in 
vitro even in the presence of an excess 

of soluble transcription factors and 
RNA polymerase III sufficient to be 
capable of efficiently transcribing oo- 
cyte-type 5S RNA genes in either high- 
molecular-weight genomic DNA or in 
recombinant plasmids. Activation of 
oocyte-type 5S RNA genes occurs 
upon washing chromatin with 0.6 M 
NaCl and transcribing it in either 
somatic or oocyte nuclear extracts 
(Fig. 62). 

Some insight into the mechanism by 
which oocyte-type 5S RNA genes are 
activated is provided by transcription 
of native and salt-washed erythrocyte 

2 3 


Fig. 62. Transcription of chromatin. 1 /xg of 
intact chromatin was prepared from Xenopus 
tissue culture cells and added to 50 /d of oocyte 
nuclear extract. Lane (1), chromatin prepared in 
low salt; lane (2), chromatin washed with 0.35 M 
NaCl; lane (3), chromatin washed with 2 M NaCl 
(the same pattern was observed when chromatin 
was washed with 0.6 M NaCl); lane (4), 5S and 45 
RNA markers prepared from transcription of 
pXbsl and pXltDNA met . Arrow indicates posi- 
tion of 5S RNA. 



chromatin in oocyte nuclear extracts 
from which the 5S-specific transcrip- 
tion factor is removed with antibodies 
raised against it. Native chromatin 
supports faithful transcription of so- 
matic-type but not oocyte- type 5S 
RNA in both complete and factor-de- 
pleted oocyte nuclear extracts. The ef- 
ficiency of somatic 5S RNA synthesis 
is unaltered by the removal of this 
transcription factor. By contrast, both 
oocyte and somatic 5S RNAs are tran- 
scribed from naked genomic DNA puri- 
fied from erythrocytes. However, 5S 
RNA synthesis is terminated in the ab- 
sence of the exogenous 5S-specific 
transcription factor, since the 5S RNA 
genes have no preexisting factors com- 
plexed with them. Salt-washed chroma- 
tin supports transcription of both 
oocyte and somatic 5S DNAs in the 
presence of the exogenous transcrip- 
tion factor. Removal of this factor from 
the oocyte nuclear extract eliminates 
oocyte-type but not somatic-type 5S 
RNA synthesis. These results strongly 
suggest that only the endogenous so- 
matic-type 5S RNA genes are already 
complexed with a somatic 5S RNA 
transcription factor in the form of a 
stable, active transcription complex. 
The formation of a stable complex in 
vitro using cloned 5S RNA genes is 
described elsewhere. We believe that 
endogenous oocyte 5S RNA genes are 
not complexed with a transcription fac- 
tor nor are they accessible for binding 
of exogenous oocyte factor in native 
chromatin. Chromatin washed with 
0.6 M NaCl becomes altered in such 
a way that oocyte-type 5S RNA genes 
can interact with exogenous transcrip- 
tion factors and RNA polymerase III. 
These experiments suggest that the 
developmental control of 5S RNA 
genes is maintained in vitro as a prop- 
erty of the chromatin template rather 
than as the source of cell-free extract in 
which the chromatin is transcribed. 
The finding that 5S RNA genes in 
chromatin retain their developmental 
control in vitro should enable us to de- 

termine the regulatory mechanisms 
that control these genes at the molecu- 
lar level. 

5S Transcription Factors in 

Oocytes and Somatic Cells Have 

Similar Properties 

H. Pelham 

In Year Book 79 we identified the 5S 
RNA gene-specific transcription factor 
as an abundant 5S RNA-binding pro- 
tein present in immature oocytes. This 
protein has the property of binding 
specifically either to 5S RNA or to the 
control region of a 5S RNA gene. We 
now have evidence that proteins with 
similar properties are present in so- 
matic Xenopus cells and in human 
tissue culture cells. We have prepared 
extracts from each of these cell types 
that are capable of transcribing added 
genes. In both extracts, the internal 
control region is necessary and suffi- 
cient to direct specific initiation on 5S 
RNA genes. Furthermore, added 5S 
RNA specifically inhibits 5S RNA syn- 
thesis in each extract, just as it does in 
oocyte extracts (Year Book 79). Thus, 
it seems likely that the ability to bind 
5S RNA is a common property of the 
5S transcription factors in all verte- 
brates, and does not simply reflect the 
storage role of the factor in immature 
amphibian oocytes. The evolutionary 
conservation of this binding property 
suggests that it has an important func- 
tion, which might be to mediate feed- 
back control of 5S RNA synthesis in 
somatic cells. 

The amount of transcription factor in 
somatic cell extracts is much lower 
than in oocyte extracts, and it limits 
the rate of 5S RNA synthesis. From 
mixing experiments we estimate that a 
single mature oocyte nucleus contains 
as much factor as about one million 
somatic cells. The low level of tran- 
scription factor in somatic cells is of 
course consistent with a feedback inhi- 
bition model. However, if feedback in- 



hibition does occur it must be a fairly 
long-term control, because pre-formed 
transcription complexes are rather re- 
sistant to inhibition by excess 5S RNA 
(Year Book 79). 

We have raised antibodies to the 
purified oocyte factor. When added to 
the oocyte nuclear extract, these anti- 
bodies inhibit transcription only of the 
5S RNA genes transcribed by RNA 
polymerase III, confirming the spe- 
cificity of the factor. The antibodies 
have a similar effect on extracts of 
Xenopus somatic cells, implying that 
the factor in these cells is antigenicaUy 
related to the oocyte factor. To deter- 
mine if it is actually the same protein, 
we used so-called "Western blots," in 
which proteins are separated by gel 
electrophoresis, transferred to nitro- 
cellulose, and incubated with anti- 
bodies and 125 I-labeled staphylococcal 
protein A. This technique revealed two 
proteins in tissue culture and liver cells 
that react with the antibodies. One of 
these appears to be identical to the oo- 
cyte factor but is present at extremely 
low levels (about 1000 molecules/cell). 
The other is slightly larger, is about 
ten times more abundant, and, like the 
oocyte protein, binds tightly to heparin- 
agarose. We suspect that this larger 
protein is a transcription factor, be- 
cause there seems to be too little of the 
oocyte-type factor to account for the 
observed rate of 5S RNA synthesis in 
the somatic-cell extracts. 

The presence of a slightly different 
factor in somatic cells cannot by itself 
account for the selective transcription 
of somatic-type 5S RNA genes in these 
cells. The somatic extracts show no 
greater preference for somatic genes 
than do the oocyte extracts, whether 
the genes are added in the form of plas- 
mids or as total genomic DNA. 

The antibodies can also be used to fol- 
low the fate of newly synthesized 5S 
RNA in the oocyte nuclear extract. 
With either oocyte or somatic 5S genes, 
some of the transcripts bind to the 
endogenous transcription factor. This 

can be demonstrated by immunoprecip- 
itation of the factor at the end of a tran- 
scription reaction (Fig. 63). The read- 
through transcripts containing extra 
bases at their 3' ends are also precip- 
itable. Interestingly, transcripts from 
the oocyte pseudogene or from altered 

12 3 4 

12 3 4 


Fig. 63. The precipitation of newly synthe- 
sized 5S RNA by antibody to transcription fac- 
tor. Plasmids were transcribed for 1 hour in the 
oocyte nuclear extract. Sepharose beads to 
which antifactor IgG had been coupled were 
then added to the mix, and subsequently pel- 
leted and washed. RNA was prepared from pro- 
teins of the supernate (Sup) and the pellet (Ppt) 
and analyzed by gel electrophoresis. An auto- 
radiogram of the gel is shown. Plasmids con- 
tained 5S genes of the following types: (1) Xbs, 
(2) Xbs +20 maxigene, (3) Xlt, (4) Xlo 



somatic genes such as the "+20 maxi- 
gene" (Year Book 78) do not form 
stable complexes with the factor (Fig. 
63). Therefore, an interaction with 5S 
RNA does not appear to be an essential 
step in transcription. This result also 
indicates that binding of the factor to 
the transcript does not merely involve 
recognition of the sequence of the con- 
trol region in RNA form, since the pseu- 
dogene and maxigene transcripts con- 
tain this sequence yet are not bound. 
Rather, the factor must contain differ- 
ent binding sites for DNA and RNA. 

The inability of pseudogene tran- 
scripts to bind to the protein could 
explain why these transcripts do not 
accumulate in oocytes; unbound mole- 
cules are presumably susceptible to 
degradation. On the other hand, the 
readthrough transcripts of the oocyte 
genes may be bound and stabilized in 
vivo, and the extra bases subsequently 
processed off. 

Regulation of Synthesis of the 
5S RNA Transcription Factor 

/. J. Jackson 

Transcription of both oocyte and 
somatic 5S RNA genes in vitro requires 
a protein factor that binds to the con- 
trol region within the genes. The factor 
has been isolated from oocytes, and an 
antigenically related protein has been 
found in somatic cells. Regulation of 
the synthesis of this transcription fac- 
tor is crucial to the early development 
of the oocyte. We have begun to ex- 
amine the synthesis of this protein in 
immature ovaries. 

Previous work has shown this pro- 
tein to be one of the most abundant 
cytoplasmic proteins in previtellogenic 
ovaries. It makes up approximately 
15% of the steady-state protein. Incu- 
bation of whole, excised ovaries in buf- 
fer supplemented with 35 S-methionine 
labels proteins synthesized de novo 
during the incubation period. Specific 
antisera have been used, in combina- 

tion with Staphylococcus protein A 
linked to Sepharose beads, to precipi- 
tate the factor from the newly synthe- 
sized proteins. This confirms that the 
ovary is the site of synthesis of the fac- 
tor, and demonstrates that synthesis 
of the factor comprises about 5% of 
total protein synthesis during the 
labeling period. The difference between 
this protein's rate of synthesis and its 
steady-state level indicates that this 
protein probably has a turnover rate 
slower than average for most ovary 
proteins. This characteristic is one ex- 
pected of a storage protein. 

The technique has been applied to oo- 
cytes at other developmental stages. 
When ovaries from very young frogs 
(only a few months post-metamorpho- 
sis) were examined, synthesis of the 
factor was detected, but the rate of 
synthesis was much less than in more 
developed ovaries. No synthesis of the 
factor could be detected in mature oo- 

Messenger RNA (poly A-containing 
RNA) has now been prepared from 
ovaries actively synthesizing the fac- 
tor, and assays are being developed to 
identify the factor among the products 
of in vitro translation of the mRNA. 
These assays will be used to identify 
clones containing cDNA complemen- 
tary to the factor mRNA in a library of 
cloned ct)NA sequences from ovary. 
The availability of a nucleic acid hy- 
bridization probe will permit the isola- 
tion of the chromosomal gene(s) coding 
for the different factors present in so- 
matic and ovarian cells. We can then 
study the developmental regulation of 
the factor gene(s). 

Establishment of the 

Developmental Control of 

5S RNA Genes in 

Xenopus Embryos 

D. D. Brown, M. Wormington, and E. Jordan 

With the development of more sensi- 
tive techniques for the analysis of 5S 



RNA, we have been able to determine 
when the control of these genes is 
established in early embryos. 5S RNA 
synthesized upon injection of 32 P-ribo- 
nucleoside triphosphates into oocyte 
nuclei or upon incubation of radioac- 
tive precursors with manually isolated 
germinal vesicles is predominantly 
oocyte-type. Progesterone treatment 
of mature oocytes results in germinal- 
vesicle breakdown and in shut-down of 
all endogenous 5S RNA synthesis. 
Plasmids containing 5S RNA, tRNA, 
or adenovirus VA RNA genes that 
were injected into oocyte nuclei and 
transcribed actively become transcrip- 
tionally inactive after progesterone- 
induced meiosis. 

Two-cell-stage embryos were in- 
jected with radioactive precursors. The 
5S RNA synthesized at later stages 
was gel purified, digested with ribonu- 
clease Tl, and identified by oligonu- 
cleotide fingerprinting. No detectable 
5S RNA synthesis was observed until 

late-blastula stage, just prior to the on- 
set of gastrulation, at which time both 
oocyte-type and somatic 5S RNA 
began to be synthesized in about equal 
amounts. By late gastrulation, mainly 
somatic 5S RNA was synthesized. 

It was of interest to determine if the 
transient activation of oocyte-type 5S 
RNA genes was due solely to transcrip- 
tion of the maternal 5S RNA genes. 
These genes could have remained com- 
plexed with the oocyte transcription 
factor following oogenesis, and this 
could account for their activation. To 
test this hypothesis, interspecies hy- 
brids between X. laevis and X. borealis 
were bred in order to distinguish the 
oocyte and somatic 5S RNAs from 
each parent. Our results clearly indi- 
cate that both paternal and maternal 
oocyte-type and somatic 5S RNA 
genes are transcribed at late blastula. 
Thus, the developmental control of 5S 
RNA synthesis occurs within a narrow 
developmental period. 


Ronan O'Rahilly 

The embryological collections, both 
human and comparative, are housed at 
the University of California, Davis, 
and all inquiries, as well as requests for 
permission for publication, should be 
addressed to Professor R. O'Rahilly, 
Carnegie Laboratories of Embryology, 
University of California, Davis, Cali- 
fornia 95616. 

Developmental Stages in 
Human Embryos 

A century after Wilhelm His, Sr., 
produced his Anatomie menschlicher 
Embryonen, and 70 years after Keibel 
pleaded for "an account of the develop- 
ment of the human body, based through- 
out on human material," we have now 
"come so much nearer the goal." The 
acquisition of very early human em- 

bryos (notably through the efforts of 
Hertig and Rock) and the staging of 
the human embryo (first by Mall and 
subsequently by Streeter) have brought 
precision to the timing and sequence of 
developmental events. Occasional for- 
ays into the histochemistry and elec- 
tron microscopy of staged human em- 
bryos have already been made, but it is 
still far from generally appreciated 
that much more work remains to be 
done with the light microscope before 
our knowledge of embryonic and fetal 
anatomy will be at all comparable to 
that of the adult organism. In attempt- 
ing to bridge these gaps, the Carnegie 
Collection— that embryological "Bu- 
reau of Standards," to use George 
Corner's judicious comparison— has 
been and continues to be of paramount 
importance. Moreover, as exemplified 



by certain studies mentioned below, we 
have now reached a point where an idea 
of the variation found within a single 
developmental stage can be assessed. 

The revision of stages 10 through 23 
is being pursued by O'Rahilly and 
M tiller, who are concentrating at pres- 
ent on the nervous system and its 
skeletal coverings. For this purpose, 
stages 8 and 9 (Fig. 64) are also being 

The computerized catalog of Car- 
negie specimens, established by Dr. 
Alexander Barry, which already in- 
cluded the Carnegie Collection, the 
Blechschmidt Collection, and the col- 
lections of the California Primate Re- 
search Center (Davis), the University 
of Washington, the University of Ala- 
bama, and the University of Michigan, 
now also includes the collection at 
Kyoto University. 

Development of the 
Nervous System 

The project on the developmental neu- 
robiology of primates has been fur- 
thered by a study of the very early nerv- 
ous system in staged insectivore and 
primate embryos (stages 11-13), pub- 
lished recently by M tiller and O'Rahilly. 
The early development of the neural 
tube is basically similar in all the 
species investigated, but differences in 
detail are noticeable. The mesencepha- 
lic flexure serves in all cases as a land- 
mark for Malpighi's tripartite subdivi- 
sion of the brain. The nonhuman embryos 
seem to show a little more variation 
than the human in the closure of the 
neuropores in relation to somitic count. 
With the exception of the later-appear- 
ing terminal vomeronasal component, 
all major portions of the neural crest as 
classified by O'Rahilly (1965) are repre- 
sented in both the nonhuman and the 
human embryos studied. No crest is 
present at the level of rhombomere 1, 
nor at rhombomere 3 except in the pla- 
tyrrhines and some human embryos, 
nor at rhombomere 5 except in certain 

human specimens. Indications of the 
initial forking of the trigeminal gang- 
lion into its primary divisions seem to 
vary with species. Ganglionic contribu- 
tions from head ectoderm (epipharyn- 
geal placodes) were found in Cebus 
apella. In general, the motor compo- 
nents of nerves are identifiable before 
the sensory elements. Nerve fibers 
were first observed in the human at 
stage 13 in some of the cranial nerves 
and in the ventral roots of the spinal 

Stage 8 (18 days), when the brain 
first appears in the human, has been re- 
investigated in 1 1 embryos. This is the 
first study in which it has been pos- 
sible to examine more than two 
specimens at a given stage, and it is 
the first account of embryos of ex- 
cellent histological quality at stage 8. 
The neural groove, which is the first 
morphological manifestation of the 
nervous system, is shown to be present 
in one-quarter of embryos at this im- 
portant stage. Correlations with other 
developmental features suggest that 
the neural groove is seen only when a 
certain degree of size and maturity has 
been reached. 

Collaboration with Professor W. 
Wozniak (Poznah) continues, and 
an electron microscopic study of the 
developing human spinal cord has been 

Professor Jean Bossy (Nimes) stud- 
ied the collection of staged embryos 
and published articles on the develop- 
ment of the olfactory and related struc- 
tures, the dorsal rami of the spinal 
nerves, and the embryonic appearance 
of the nipple. 

An account of the growth of the brain 
during the embryonic period proper by 
Drs. Mary E. Desmond (Villanova) and 
R. O'Rahilly is awaiting publication. 

The Developing Skeletal System 

A detailed account of the chondro- 
cranium at stage 23 (8 weeks), with par- 
ticular reference to the nervous sys- 



tem, has been published recently by 
Miiller and O'Rahilly. The authors are 
particularly pleased that their graphic 
and Perspektomat reconstructions 

have met with the approval of Mr. Os- 
borne O. Heard, who, over some 42 
years, made practically all the solid- 
state reconstructions in the Carnegie 




Fig. 64. Graphic reconstruction made by F. Miiller with the aid of the newly acquired 
Perspektomat P-40 apparatus. The figure shows the dorsal surface of a human embryo (No. 7650) of 
stage 9, a stage that is very rarely found. The embryo, which was donated to the Carnegie Collection 
by Dr. Arthur T. Hertig, is 2-3 mm in length and possesses 2-3 pairs of somites. The neural groove 
can be seen in the upper part of the figure, and the allantoic diverticulum (cut) below. The obliquity 
of the histological sections prevents reconstruction by the older techniques. 



Collection, many of which are illus- 
trated in practically every textbook of 
embryology. Ten embryos at state 23 
were studied; variations are minor and 
appear not to be closely related to each 
other. Hence, it was not possible to ar- 
range the embryos in a linear develop- 
mental series based on their level of 
cranial morphogenesis. In general, the 
degree of cranial development at stage 
23 is intermediate between that at 
stage 20 (the only other skull so far 
studied adequately in a staged embryo) 
and that of fetuses of 40 and 43 mm. 
Ossification is beginning in the tectum 
posterius, and most of the "mem- 
brane" components show osseous 

Detailed studies of the embryonic ver- 
tebral column are being continued by 
Drs. O'Rahilly, Muller, and David B. 
Meyer (Detroit). An account of the col- 
umn at stage 23 (8 weeks), based on 
precise reconstructions, has appeared. 
Scarcely any illustrations of either the 
cartilaginous column as a whole or of 
individual vertebrae in the embryo 
have been available previously. This is 
all the more surprising in view of the 
frequency and importance of such 
anomalies as spina bifida. It was found 
that the posterior view of the normal 
column at stage 23 bears a striking 
resemblance to a total rhachischisis in 
the newborn. Stage 23 shows typically 
33 or 34 cartilaginous vertebrae ar- 
ranged in flexion and approximately 
20-33 mm in total length. A typical 
vertebra consists of a centrum that is 
continuous with two neural processes. 
Pedicles, and articular and transverse 
processes, but no spinous processes, 

are identifiable. The variations within 
stage 23 were mostly minor, e.g., the 
number (4-6) of coccygeal elements 
and the extent of the dorsal growth of 
the neural processes. The union of the 
laminae, the onset of ossification, and 
the appearance of articular cavities are 
all characteristic of the early fetal 
period. Normal and abnormal fetal col- 
umns are being studied by Dr. Daniel 
R. Benson (Davis). 

Other Organs 

An account of the musculature and 
innervation of the larynx at stage 23 
by Drs. Muller, O'Rahilly, and John 
A. Tucker (Philadelphia) is awaiting 

Dr. Pieter deVries (Davis) published 
some of his ongoing studies of the de- 
veloping heart, with particular refer- 
ence to the truncus and infundibulum. 
Drs. Grover M. Hutchins and G. Wil- 
liam Moore (Baltimore) continued their 
work on the heart and published on the 
interventricular septum and supraven- 
tricular crest. 


Visitors who studied the collection 
included Drs. Gerald R. Cunha (Den- 
ver), Mary E. Desmond (Villanova, 
Pennsylvania), Virginia E. Diewert 
(Vancouver), R. Leon Hughes (Bris- 
bane, Australia), Grover M. Hutchins 
(Baltimore), David B. Meyer (Detroit), 
G. William Moore (Baltimore), William 
Murphy (Alameda, California), John A. 
Tucker (Philadelphia), and Doris B. 
Wilson (San Diego). 


During the year, Staff Members par- 
ticipated as chairpersons and speakers 
at the following conferences: Gordon 
Research Conferences on Lysosomes, 
Plant Molecular Biology, Extrachro- 

mosomal Elements, Plant Tissue Cul- 
ture, Nucleic Acids, Animal Cells and 
Viruses, Regulation, Lipid Metabo- 
lism, and Immunology; Cold Spring 
Harbor Laboratory Conferences on C. 



elegans, Neurobiology, herpes viruses, 
and "Monoclonal antibodies to neural 
antigens;" Society for Neuroscience; 
Dahlem Conference on "Neuron-Glial 
Interactions and Multiple Sclerosis," 
Berlin; Taniguchi Symposium, Lake 
Biwa, Japan; Symposia on Cell Inter- 
actions, University of Minnesota 
School of Medicine; Muscular Dystro- 
phy Association International Con- 
ference, Miami; ICN-UCLA Symposia 
on Developmental Biology and on Im- 
munoglobulin Idiotypes, Keystone, 
Colorado; Fourth International Con- 
gress of Immunology; American Soci- 
ety for Cell Biology; New York Society 
of Electron Microscopists; Anato- 
mische Gesellschaft, Wurzburg; Feder- 
ation of European Biochemical Soci- 
ety, Edinburgh; Genetics Society of 
America; International Cell Biology 
Congress, Berlin; British Genetical So- 
ciety; National Drosophila Meeting, 
Chicago; Drosophila Molecular Biol- 
ogy Workshop, Crete; Given Institute 
Conference on Eukaryotic Genes, As- 
pen; The 9th International Conference 
on Biological Membranes, Crans. 

Lectures were given at NIH, Johns 
Hopkins University (School of Hygiene 
and Medical School), Cornell Uni- 
versity, Pennsylvania State Medical 
School (Hershey), University of Mary- 
land Medical School and Baltimore 
County campus, University of Pennsyl- 
vania, National Headquarters of Amer- 
ican Red Cross, University of Dela- 
ware, Tufts University, Washington 
University, Michigan State Univer- 
sity, Fox Chase Institute for Cancer 
Research, Georgetown University, Cal- 
ifornia Institute of Technology, Uni- 
versity of California at San Diego, at 
Berkeley, and at Davis, Stanford Uni- 
versity Medical School, Hopkins Ma- 
rine Biology Station, Harvard Univer- 
sity and Harvard Medical School, Uni- 
versity of Virginia, Duke University, 
Marine Biological Laboratories at 
Woods Hole, Wesleyan University, 
Case Western Reserve University, 
Rockefeller University, Northwestern 

University Medical School, S.U.N.Y. at 
Stony Brook, at Brooklyn, and at Buf- 
falo, University of Chicago, Imperial 
Cancer Research Fund (London), MRC 
Laboratory of Molecular Biology (Cam- 
bridge), Massachusetts Institute of 
Technology, Yale University, New Jer- 
sey College of Medicine, Albert Ein- 
stein College of Medicine, University of 
North Carolina, Columbia University, 
Emory University, University of Ro- 
chester, Frederick Cancer Research 
Center, Osaka University, Kyoto Uni- 
versity, Rutgers University School of 
Medicine, Colorado State University, 
Mayo Clinic, the Wellcome Lecture at 
Louisiana State University, and the 
Harvey Lecture, New York City. 

Members of the Staff served on the 
Science Advisory Committee, Office of 
Technology Assessment— Congress of 
the United States, and National Insti- 
tutes of Health Recombinant DNA Ad- 
visory Committee; the editorial boards 
of Gene, Neuroscience, Journal of 
Neuroscience Methods, Developmental 
Neuroscience, and Developmental 
Brain Research. Staff Members were 
members of the Cellular Physiology 
Review Group, a special study section 
of General Medical Sciences, and the 
Molecular Cytology Study Section, all 
of NIH. Additional consulting ac- 
tivities included scientific advising for 
the Searle Scholars Program, member- 
ships on the visiting committees of the 
Roche Institute of Molecular Biology 
(Chairman); the Biology (Chairman) and 
Molecular Biology Department of 
Harvard University; the Centre de 
Genetique Moleculaire of the CNRS, 
Gif-sur-Yvette; the Board of Scientific 
Counselors of the National Institute of 
Child Health and Human Develop- 
ment; and the Department of Biochem- 
istry, Princeton University. 


The Department offers at least two 
seminars each week, one of them usu- 



ally presented by a scientist from out- 
side the Department. This year 15 in- 
vited speakers from the Baltimore 
scientific community lectured. The 
fourth annual Mini-Symposium was 
entitled "Genetic and Biochemical Ap- 
proaches to Cellular Interactions." The 
four principal speakers were Julius Ax- 
elrod, National Institute of Mental 
Health; Victor D. Vacquier, Scripps In- 
stitute of Oceanography; Ursula Good- 
ehough, Washington University; and 
Richard Lerner, Scripps Clinic and Re- 
search Foundation. In addition to this 
Symposium, the following visitors 
spoke: Hal Gainer, Igor Dawid, Dean 
Hamer, George Khoury, Barbara Bat- 
telle, W. French Anderson, Robert 
Rohwer, and Steve O'Brien, all from 
the National Institutes of Health; C. 
Benyajati, Frederick Cancer Research 
Center; Andrew Lamer, Sandy Bern- 
stein, University of Virginia; Cathy 
French, University of North Carolina; 

Nadine Weich, Sinai Medical School; 
Samit Adhya, Albert Einstein College 
of Medicine; Mary Schuler, Cornell 
University; David Housman, Massa- 
chusetts Institute of Technology; Eric 
Frank and Carl Wu, Harvard Univer- 
sity; Joe Gall, Yale University; Laura 
Kalfayan, Brandeis University; Sam 
Rose, University of Illinois; Richard 
Sleight, Purdue University; Bill En- 
gels, Mark Emmerling, and Carl John- 
son, University of Wisconsin; Tom 
Chech and Steve Roper, University of 
Colorado; Rick Dahlquist, University 
of Oregon; Leonard Rabinow, Univer- 
sity of Utah; Mel Green and Kenneth 
Longmuir, University of California; 
Einhard Schierenberg, Max-Planck-In- 
stitut fur Experimen telle Medizin; 
William Sharrock and Steve Anderson, 
MRC Unit; Peter Wellauer, Swiss In- 
stitute for Experimental Cancer Re- 


Argon, Y., and S. Ward, Caenorhadbitis ele- 
gans fertilization-defective mutants 
with abnormal sperm, Genetics, 96, 413- 
433, 1981. 

Bogenhagen, D. F., and D. D. Brown, Nu- 
cleotide sequences in Xenopus 5S DNA 
required for transcription termination, 
Cell, 24, 261-270, 1981. 

Bogenhagen, D. F., see Wormington, W. M. 

Bossy, J., Development of olfactory and 
related structures in staged human em- 
bryos, Anat. Embryol, 161, 225-236, 

Bossy, J., Developpement des branches 
dorsales des nerfs spinaux durant la pe- 
riode embryonnaire vraie, Bull Ass. 
Anat, 64, 199-206, 1980. 

Bossy, J., Developpement sequentiel 
de la glande mammaire primitive pen- 
dant la periode embryonnaire vraie, Seno- 
logia, 5, 325-328, 1980. 

Boving, B. G., Vascular relations of im- 
planting human conceptuses, Anat. Rec, 
199, 33A-34A, 1981. 

Brorein, W. J., see Dunsmuir, P., and 
Rubin, G. M. 

Brown, D. D., Gene expression in eukary- 
otes, Science, 211, 667-674, 1981. 

Brown, D. D., see Bogenhagen, D. F., 
Pelham, H. R. B., Sakonju, S., and Worm- 
ington, W. M. 

Cebra, J. J., P. J. Gearhart, J. F. Halsey, J. 
L. Hurwitz, and R. C. Shahin. Role of en- 
vironmental antigens in the ontogeny of 
the secretory immune response, J. Retic- 
uloendothel. Soc, 28, 61S-71S, 1980. 

Cebra, J. J., see Gearhart, P. J. 

Davignon, R. W., R. M. Parker, and A. G. 
Hendrickx, Staging of the early embry- 
onic brain in the baboon (Papio cynoceph- 
alus) and rhesus monkey (Macaca mul- 



atta), Anat. Embryol, 159, 317-334, 

Devreotes, P. N., see Fambrough, D. M. 

deVries, P. A., Evolution of precardiac and 
splanchnic mesoderm in relationship to 
the infundibulum and truncus, Perspect. 
Cardiovasc. Res., 5, 31-48, 1981. 

Diewert, V. M. Active contributions of dif- 
ferential craniofacial growth to second- 
ary palate development in man, Anat. 
Rec, 199, 69A-70A, 1981. 

Douglas, R., see Kehry, M., and Gearhart, 

Dunsmuir, P., W. J. Brorein, Jr., M. A. 
Simon, and G. M. Rubin, Insertion of the 
Drosophila transposable element copia 
generates a 5 base pair duplication, Cell, 
21, 575-579, 1980. 

Dunsmuir, P., see Levis, R., and Rubin, 
G. M. 

Ewald, S., see Kehry, M. 

Fambrough, D. M., Some thoughts on solv- 
ing the problem of synaptogenesis, in On- 
togenesis and Functional Mechanisms of 
Peripheral Synapses, J. Taxi, ed., 
Elsevier/North Holland, Amsterdam, pp. 
355-363, 1980. 

Fambrough, D. M., D. C. Linden, P. N. 
Devreotes, J. M. Gardner, and R. L. 
Rotundo, On the regulation of acetylcho- 
line receptors in skeletal muscle, in Nerve 
Cells, Transmitters and Behavior, Scrip- 
ta Varia 45, Pontifical Academy of Sci- 
ences, R. Levi-Montalcini, ed., Elsevier, 
Amsterdam, pp. 485-509, 1980. 

Fambrough, D. M., see Kehry, M., and 
Rotundo, R. L. 

Flavell, A. J., S. W. Ruby, J. J. Toole, B. E. 
Roberts, and G. M. Rubin, Translation 
and developmental regulation of RNA en- 
coded by the eukaryotic transposable ele- 
ment copia, Proc. Nat. Acad. Sci. USA, 77, 
7107-7111, 1980. 

Flavell, A. J., see Rubin, G. M. 

Gardner, E., see O'Rahilly, R. 

Gardner, J. M., see Fambrough, D. M. 

Gavis, E. R., see McKnight, S. L. 

Gearhart, P. J., J. L. Hurwitz, and J. J. 
Cebra, Successive switching of antibody 

isotypes expressed within the line of a B- 
cell clone, Proc Nat. Acad. Sci. USA, 77, 
5425-5428, 1980. 

Gearhart, P. J., N. D. Johnson, R. Douglas, 
and L. Hood, IgG antibodies to phos- 
phorylcholine exhibit more diversity 
than their IgM counterparts, Nature, 
291, 29-34, 1981. 

Gearhart, P. J., see Cebra, J. J. 

Halsey, J. F., see Cebra, J. J. 

Hendrickx, A. G., see Davignon, R. W. 

Hipskind, R. A., see McKnight, S. L. 

Hoekstra, D., see Struck, D. K. 

Hood, L., see Gearhart, L. and Kehry, M. 

Hurwitz, J. L., see Cebra, J. J., and 
Gearhart, P. J. 

Hutchins, G. M., see McBride, R. E., 
Meredith, M. A., and Moore, G. W. 

Johnson, N. D., see Gearhart, P. J. 

Jordan, E., see Wormington, M. W. 

Kehry, M., S. Ewald, R. Douglas, C. Sibley, 
W. Raschke, D. M. Fambrough, and L. 
Hood, The immunoglobulin /x chains of 
membrane-bound and secreted IgM mole- 
cules differ in their C-terminal segments, 
Cell, 21, 393-406, 1980. 

Levis, R., P. Dunsmuir, and G. M. Rubin, 
Terminal repeats of the Drosophila trans- 
posable element copia: nucleotide se- 
quence and genomic organization, Cell, 
21, 581-588, 1980. 

Levis, R., see Rubin, G. M. 

Linden, D. C, see Fambrough, D. M. 

Martin, O. C, see Pagano, R. E. 

McBride, R. E., G. W. Moore, and G. M. 
Hutchins, Development of the outflow 
tract and closure of the interventricular 
septum in the normal human heart, 
Amer. J. Anat, 160, 309-331, 1981. 

McKnight, S. L., R. A. Hipskind, and R. H. 
Reeder, Ultrastructural Analysis of Ribo- 
somal Transcriptional in vitro, J. Biol. 
Chem., 255, 7907-7911, 1980. 

McKnight, S. L., and E. R. Gavis, Expres- 
sion of the Herpes simplex virus thy- 
midine kinase in Xenopus laevis oocytes: 
an assay for the study of deletion mu- 



tants constructed in vitro, Nucl. Acids. 
Res., 8, 5931-5948, 1980. 

McKnight, S. L., The nucleotide sequence 
and transcript map of the Herpes sim- 
plex virus thymidine kinase gene, Nucl. 
Acids Res., 8, 5949-5964, 1980. 

Meredith, M. A., G. M. Hutchins, and G. 
W. Moore, Role of the left interventricu- 
lar sulcus in formation of the interven- 
tricular septum and crista supraventricu- 
laris in normal human cardiogenesis, 
Anat. Rec, 194, 417-428, 1979. 

Meyer, D. B., see O'Rahilly, R. 

Moore, G. W., G. M. Hutchins, and R. 
O'Rahilly, The estimated age of staged 
human embryos and early fetuses, Amer. 
J. Obstet. Gynecol., 139, 500-506, 1981. 

Moore, G. W., see McBride, R. E., and 
Meredith, M. A. 

Muller, F., and R. O'Rahilly, The early de- 
velopment of the nervous system in 
staged insectivore and primate embryos, 
J. Comp. Neurol., 193, 741-751, 1980. 

Muller, F., and R. O'Rahilly, The human 
chondrocranium at the end of the em- 
bryonic period proper, with particular ref- 
erence to the nervous system, Amer. J. 
Anat., 159, 33-58, 1980. 

Muller, F., see O'Rahilly, R. 

Muller, K. J., and S. A. Scott, Transmission 
at a "direct" electrical connexion me- 
diated by an interneurone in the leech, J. 
Physiol. (London), 311, 565-583, 1981. 

Muller, K. J., see Scott, S. A. 

Nelson, G. A., and S. Ward, Amoeboid mo- 
tility and actin in Ascaris lumbricoides 
sperm, Exp. Cell Res., 131, 149-160, 

Ng, S.-Y., see Sakonju, S. 

Nichols, J. W., and R. E. Pagano, Kinetics 
of soluble lipid monomer diffusion be- 
tween vesicles, Biochemistry, 20, 2783- 
2789, 1981. 

Olszewska, B., see Wozniak, W. 

O'Rahilly, R., Le developpement prenatal, 
C est Pour Quand 3, 16-19, 1980. 

O'Rahilly, R., and E. Gardner, Initial devel- 
opment of the human brain, Bibl. Anat, 
19, 140-141, 1980. 

O'Rahilly, R., and F. Muller, Embryonic 
characteristics associated with the ap- 
pearance of the neural groove, Anat. 
Rec, 199, 187A-188A, 1981. 

O'Rahilly, R., F. Muller, and D. B. Meyer, 
The human vertebral column at the end 
of the embryonic period proper. 1. The 
column as a whole, J. Anat, 131, 565- 
575, 1980. 

O'Rahilly, R., see Moore, G. W., Muller, F., 
and Wozniak, W. 

Pagano, R. E., O. C. Martin, A. J. Schroit, 
and D. K. Struck, Formation of asymmet- 
ric vesicles via spontaneous transfer of 
fluorescent lipid analogues between ves- 
icle population, Biochemistry, 20, 4920- 
4927, 1981. 

Pagano, R. E., see Nichols, J. W., Schroit, 
A. J., and Struck, D. K. 

Parker, R. M., see Davignon, R. W. 

Pelham, H. R. B., and D. D. Brown, A spe- 
ific transcription factor that can bind 
either the 5S RNA gene or 5S RNA, Proc. 
Nat. Acad. Sci. USA, 77, 4170-4174, 

Pelham, H. R. B., W. M. Wormington, and 
D. D. Brown, Related 5S RNA transcrip- 
tion factors in Xenopus oocytes and 
somatic cells, Proc. Nat. Acad. Sci. USA, 
78, 1760-1764, 1981. 

Raschke, W., see Kehry, M. 

Reeder, R. H., see McKnight, S. L. 

Roberts, B. E., see Flavell, A. J. 

Roeder, R. G., see Sakonju, S. 

Rotundo, R. L., and D. M. Fambrough, 
Synthesis, transport and fate of acetyl- 
cholinesterase in cultured chick embryo 
muscle cells, Cell, 22, 583-594, 1980. 

Rotundo, R. L., and D. M. Fambrough, Se- 
cretion of acetylcholinesterase; relation 
to acetylcholine receptor metabolism, 
Cell, 22, 595-602, 1980. 

Rotundo, R. L., see Fambrough, D. M. 

Rubin, G. M., W. J. Brorein, Jr., P. 
Dunsmuir, A. J. Flavell, R. Levis, E. 
Strobel, J. J. Toole, and E. Young, Copia- 
like transposable elements in the Dro- 
sophila genome, Cold Spring Harbor Symp. 
Quant. Biol., 45, 619-628, 1981. 



Rubin, G. M., see Dunsmuir, P., Flavell, A. 
J., and Levis, R. 

Ruby, S. W M see Flavell, A. J. 

Sakonju, S., D. D. Brown, D. Engelke, S.-Y. 
Ng, B. S. Shastry, and R. G. Roeder, The 
binding of a transcription factor to dele- 
tion mutants of a 5S ribosomal RNA 
gene, Cell, 23, 665-669, 1981. 

Schroit, A. J., and R. E. Pagano, Capping 
of a phospholipid analog in the plasma 
membrane of lymphocytes, Cell, 23, 
105-112, 1981. 

Scott, S. A., and K. J. Muller, Synapse 
regeneration and signals for directed ax- 
onal growth in the central nervous sys- 
tem of the leech, Develop. Biol., 80, 
345-363, 1980. 

Scott, S. A., see Muller, K. J. 

Shastry, B. S., see Sakonju, S. 

Sibley, C, see Kehry, M. 

Simon, M. A., see Dunsmuir, P. 

Strobel, R., see Rubin, G. M. 

Struck, D. K., D. Hoekstra, and R. E. 
Pagano, Use of resonance energy transfer 
to monitor membrane fusion, Biochemis- 
try, 20, 4093-4099, 1981. 

Toole, J. J., see Flavell, A. J., and Rubin, 
G. M. 

Ward, S., see Argon, Y., and Nelson, G. A. 

Wormington, W. M., D. F. Bogenhagen, E. 
Jordan, and D. D. Brown, A quantitative 
assay for Xenopus 5S RNA gene tran- 
scripion in vitro, Cell, 24, 809-817, 1981. 

Wormington, W. M., see Pelham, H. R. B. 

Wozniak, W., R. O'Rahilly, and B. Olszew- 
ska, The fine structure of the spinal cord 
in human embryos and early fetuses, 
J. Hirnforsch. 21, 101-124, 1980. 

Young, E., see Rubin, G. M. 


Year Ended June 30, 1981 
(Including those whose services ended during the year) 

Research Staff 

Donald D. Brown, Director 
Douglas M. Fambrough 
Nina V. Fedoroff 
Kenneth J. Muller 
Richard E. Pagano 
Gerald M. Rubin 1 
Allan Spradling 
Samuel Ward 

Staff Associates 

Patricia Gearhart 
Steven L. McKnight 2 
Richard Rotundo 

^rom September 1, 1980 
2 To December 31, 1980 

Research Associates (Extramural) 

Bent Boving, Detroit, Michigan 
Igor B. Dawid, Bethesda, Maryland 
Robert L. DeHaan, Atlanta, Georgia 
Arthur T. Hertig, Boston, Massachu- 
Irwin R. Konigsberg, Charlottesville, 

Ronan O'Rahilly, Davis, California 
Elizabeth M. Ramsey, Washington, D.C. 
Ronald H. Reeder, Seattle, Washington 
Yoshiaki Suzuki, Okazaki City, Japan 

Postdoctoral Fellows and Grant- 
Supported Associates 

M. John Anderson, Research Associ- 
ate, Muscular Dystrophy (Fambrough) 

Ellen Bayne, Fellow of Carnegie In- 
stitution of Washington 

Daniel Bogenhagen, Fellow of Helen Hay 
Whitney Foundation 



Deborah Chaleff, 1 Fellow of the Damon 
Runyon-Walter Winchell Cancer Fund 

Matthias Chiquet, 2 Fellow of the Swiss 
National Science Foundation 

Mary Collins, 3 Fellow of the National 
Institutes of Health 

Ellen Elliott, Fellow, National Multiple 
Sclerosis Society 

Mark Emmerling, 4 Fellow of the Muscu- 
lar Dystrophy Association 

Dick Hoekstra, Fellow of the Carnegie 
Institution of Washington, and Re- 
search Associate, National Institutes 
of Health Grant (Pagano) 

Ian Jackson, 5 Fellow of the European 
Molecular Biology Organization 

Roger Karess, 6 Fellow of the National In- 
stitutes of Health 

Robert Levis, 7 Research Associate, Na- 
tional Institutes of Health Grant (Ru- 

Adrian Mason, 8 Research Associate, Na- 
tional Institutes of Health Grant (Mul- 

Sheila McCormick, Fellow of the Na- 
tional Institutes of Health 

Wylie Nichols, Fellow of the National In- 
stitutes of Health 

Kevin O'Hare, 9 Research Associate, 
American Cancer Society Grant (Rubin) 

Hugh Pelham, Research Associate, Na- 
tional Institutes of Health Grant 

Ronald Peterson, 10 Fellow of the Helen 
Hay Whitney Foundation 

Thomas Roberts, 11 Fellow of the National 
Institutes of Health 

Richard Rotundo, Fellow of the Muscular 
Dystrophy Association 

Alan Schroit, 12 Research Associate, Na- 
tional Institutes of Health Grant (Pa- 

Mavis Shure, Fellow of the Damon Ru- 
nyon-Walter Winchell Cancer Fund 

Douglas Struck, 13 Fellow of the Amer- 
ican Cancer Society 

Adina Student, 14 Fellow of the Muscular 
Dystrophy Association 

Barbara Wakimoto, 15 Fellow of the Helen 
Hay Whitney Foundation 

Eric Wakshull, 16 Fellow of the Muscular 
Dystrophy Association 

Susan Wessler, Fellow of the American 
Cancer Society 

Michael Wormington, Research Associ- 
ate, National Institutes of Health 
Grant (Brown) 


Christopher Austin, Undergraduate, 
Princeton University 

Dan Burke, Graduate, National Research 
Council of Canada 

John M. Gardner, Graduate, Johns Hop- 
kins University 

Elizabeth Gavis, Undergraduate, Cornell 

Elizabeth Liebson, Undergraduate, Wel- 
lesley College 

Benjamin Nathans, Undergraduate, Yale 

Suki Parks, Graduate, Johns Hopkins 

Shigeru Sakonju, Graduate, Johns Hop- 
kins University 

Mark Schlissel, Medical Science Traning 
Program, Johns Hopkins University 

Jennifer Schwartz, Graduate, Johns 
Hopkins University 

Visiting Investigators and Extramural 

1 From August 1, 1980 

2 From December 1, 1980 

3 From September 1, 1980 

4 From October 1, 1980 

5 From December 1, 1980 

6 From October 1, 1980 

7 From November 1, 1980 

8 From October 1, 1980 

9 From June 1, 1980 
10 To September 15, 1980 
"To January 31, 1981 
12 To August 31, 1980 

P. Bingham, Research Triangle Park, 

North Carolina 
J. Cebra, Philadelphia, Pennsylvania 
A. G. Engel, Rochester, Minnesota 
J. G. Gall, New Haven, Connecticut 
K. Longmuir, Irvine, California 
L. Mets, Cleveland, Ohio 
T. L. Rosenberry, Cleveland, Ohio 

13 To August 31, 1980 
'From July 1, 1980 

15 From May 1, 1981 
16 From August 31, 1981 



Clerical and Technical Staff 

James H. Blackwell, Custodian 
Paul Blackwell, Custodian (part-time) 
Patricia Chase, l Laboratory Helper 
William H. Duncan, Senior Technician 
Ernestine V. Flemmings, Laboratory 

Richard D. Grill, Photographer 
Virginia Hicks, Laboratory Helper 
Mary E. Hogan, Technician 
John E. Jones, Custodian 
Eddie Jordan, Senior Technician 
Nancy Jordan, Laboratory Helper (part- 
Robert Kingsbury, 2 Technician 
Catherine R. Lane, Librarian (part-time) 

Joseph Levine, 3 Technician 

Alice H. Mabin, Laboratory Helper 

Thomas F. Malooly, Business Manager 

Ona Martin, Technician 

Jeffrey Mauvais, Technician 

Thomas F. Miller, Custodian 

Christine Murphy, 4 Technician 

Nadine Nivera, Technician 

John Pazdernik, Building Engineer 

Leslie Pederson, 5 Technician 

Betty Lou Phebus, Bookkeeper/Clerk 

Susan Satchell, Secretary 

Pat Schmidt, Secretary 

Audrey Sheppard, Laboratory Helper 

Delores Somerville, Senior Technician 

Barbara Thomas, Technician 

Joe Vokroy, Machinist 

! To August 31, 1980 
2 To December 31, 1980 

3 From August 1, 1980 

4 From September 17, 1980 

5 To August 15, 1980 

Developmental Biology 
Research Group 

Pasadena, California 

and Kerckhoff Marine Laboratory, 

Corona del Mar, California 

Eric H. Davidson 
Professor, California Institute of Technology 

Roy J. Britten 
Distinguished Carnegie Senior Research Associate 


Introduction 233 

Sequence polymorphism in regions surrounding structural genes 233 

DNA sequence evolution around a sea urchin gene 234 

Distant homology or noise? 235 

Interspersed repetitive sequences in the human genome 235 

Transcription of human repeat sequences 236 

Developmental expression of sea urchin repetitive sequence families 236 

Poly (A) +RNA of amphibian oocytes 237 

Structure and fate of sea urchin maternal RNA 237 

Antibodies to proteins of the vitelline layer of sea urchin eggs 238 

Expression of cloned sea urchin genes in embryo and adult cells 238 

Sea urchin embryo specific expression of RNA sequences 239 

Regulation of prevalent mRNAs during sea urchin embryonic development 240 

Control of mRN A prevalence in the sea urchin embryo. 1. cytoplasmic RNA 240 

Control of mRNA prevalence in the sea urchin embryo. 2. nuclear RNA 241 

A transcription unit coding for multiple poly (A) + RN As in the sea urchin 241 

Transcription in isolated sea urchin nuclei 242 

The sea urchin actin gene set 242 

Expression of actin genes in the sea urchin embryo 243 

Expression of intracisternal A-type particles in early mouse embryos 244 

Publications 244 

Staff 245 

Support 246 


The work of this laboratory is con- 
cerned with the molecular biology of 
development. This has come to include 
basic studies of the organization of 
genomic DNA sequences, and of changes 
in single-copy (nonrepetitive) and re- 
peated DNA sequences both between 
individuals (DNA polymorphism) and 
in evolution. Much of the analysis of 
DNA relationships has been done with 
sea urchins, in particular with the local 
southern California purple sea urchin, 
Strongylocentrotus purpuratus. We 
keep a large stock of these sea urchins 
in running seawater tanks at the Cali- 
fornia Institute of Technology Division 
of Biology's Kerckhoff Marine Labora- 
tory at Corona del Mar, where several 
members of our laboratory are conduct- 
ing their research. 

One of the central issues currently 
addressed by molecular biologists is 
the question of how differential gene 
expression is regulated in higher ani- 
mals. In many of the experiments in 
our laboratory, we are using sea urchin 
eggs and developing sea urchin em- 

bryos as a system wherein to investi- 
gate the appearance and modulation of 
gene products, focusing on messenger 
and nuclear RNAs. Evidence is accu- 
mulating which indicates that ma- 
ternal messenger RNA, stored in the 
egg during oogenesis, shares some of 
the characteristics of mRNA precur- 
sors usually found only in the cell nu- 
cleus. The maternal RNAs include long 
molecules containing transcripts of re- 
petitive DNA sequences which do not 
code for proteins. We are using cloned 
sea urchin DNA to study the "repeat 
transcripts" as well as coding sequence 
transcripts, and their presence in egg 
and embryo RNAs. RNA from adult 
cells has been compared to embryo 
RNA in hybridizations with recombi- 
nant DNA clones derived from embryo 
messenger RNAs. These cDNA clones 
are also being hybridized with labeled 
embryo RNAs in measurements of the 
rates of synthesis (or appearance) and 
turnover of individual developmentally 
regulated RNAs. 




John W. Roberts, John W. Grula, Terry L. Thomas, Eric H. Davidson, and Roy J. Britten 

Thermal stability measurements 
have demonstrated an average 4% 
single-copy DNA sequence polymor- 
phism in the sea urchin Strongylocen- 
trotus purpuratus with a range of se- 
quence polymorphism from less than 
1% to more than 10% (Britten et al, 
1978). The blot hybridization method 
has been used to estimate the sequence 
polymorphism surrounding several S. 
purpuratus structural genes. Radioac- 
tively labeled cDNA clones from gas- 
trula poly(A) + cytoplasmic RNA were 
used as probes for hybridization to 
blots of restriction-endonuclease-di- 
gested genomic DNA from S. pur- 

puratus individuals. The sequences 
surrounding the four genes exhibited a 
range of polymorphism from less than 
1% to about 4%. Two of the genes, 
SpG19 and SpG30, appear to occur in 
regions with little or no polymorphism. 
The other two genes are located in 
more variable regions, with SpG6 
exhibiting a calculated 2% polymor- 
phism and Sp88 exhibiting about 4%. 
These results confirm the thermal 
stability measurements and, further, 
demonstrate different degrees of poly- 
morphism surrounding different genes. 
Individual genome blot hybridiza- 
tion experiments with SpG6 suggested 




patterns in the sequence polymor- 
phism. For all restriction enzymes 
tested, there appeared to be one or two 
predominant fragment lengths in the 
population, along with a series of less- 
frequent variants. This observation 
was confirmed by blot hybridization 
experiments using balanced mixtures 
of genomic DNA from 20-30 individ- 
uals from two different locations. Deep 
water (Point Loma) and intertidal 
(Laguna Beach) individuals share the 
same predominant and less-frequent 
variants. Quantitative differences in 
the frequencies of some variants, sug- 

gesting some degree of isolation be- 
tween the animals in these locations, 
cannot be excluded at this time. For 
some restriction enzymes, a series of 
fragments differing in length by 
0.2-0.5 kb were detected in the popula- 
tions, raising the possibility that small 
rearrangements contribute to the pro- 
duction of restriction fragment poly- 


Britten, R. J., A. Cetta, and E. H. Davidson. 

(1978), Cell 15: 1175-1186. 


John W. Grula, John W. Roberts, Laurence A. Lasky, Ze'ev Lev, Eric H. Davidson, 

and Roy J. Britten 

Little is known about the nature of 
DNA sequence changes underlying 
evolutionary change on the phenotypic 
level. It has been proposed that rear- 
rangements affecting gene regulation 
are more important than base substitu- 
tions (Britten and Davidson, 1971; 
Wilson et al, 1974). To explore the pre- 
valence of DNA rearrangements during 
evolution, we have examined genome re- 
gions of two sea urchin species contain- 
ing a homologous structural gene. This 
has been done by comparing recombi- 
nant lambda clones from libraries of 
DNA from Strongylocentrotus pur- 
puratus and S. franciscanus, which were 
screened with a cDNA clone made from 
S. purpuratus gastrular poly(A) + RNA. 
Several lambda clones from each 
species have been mapped by restric- 
tion enzyme analysis, and sequence 
homology between inserts has been de- 
termined by EM studies of heterodu- 
plexes and Southern blot experiments. 

The most important findings to date 
are the following. (1) If there is an inter- 
vening sequence in the S. franciscanus 
gene, it is smaller than the S. purpura- 

tus intervening sequence. (2) Most of 
the DNA flanking the 5' end of the 
genes lacks homology by the criteria of 
the methods used. (3) A moderately re- 
petitive, interspersed element close to 
the putative 5 ' end of the S. pur- 
puratus gene is absent from this posi- 
tion in S. franciscanus. However, this 
element is moderately frequent and in- 
terspersed elsewhere in the S. fran- 
ciscanus genome. This indicates that a 
rearrangement event involving an 
interspersed repetitive sequence has 
occurred near the 5 ' ends of the homol- 
ogous genes. Subclones are being con- 
structed to more precisely characterize 
this rearrangement and the other ob- 
servations by detailed restriction map- 
ping and base sequence determination. 


Britten, R. J., and E. H. Davidson. (1971), 

Quart. Rev. Biol. 46: 111-133. 
Wilson, A. C, L. R. Maxson, and V. M. Sarich. 

(1974), Proc. Nat. Acad. Sci. USA 71: 





Terrence J. Hall and Roy J. Britten 

It is possible to examine imperfectly 
matched sequence relationships, if hy- 
bridization of DNA fragments is car- 
ried out at open criteria of accuracy 
(low-temperature or high-salt condi- 
tions). However, accidental sequence 
matches may give false signals or 
noise. Deininger and Schmidt (1979) 
observed that when such conditions 
(50°C, 0.2-0.5 M sodium ion concentra- 
tion) are used for human DNA, an addi- 
tional set of low-frequency sequences 
appears to form duplexes that bind to 
hydroxylapatite. We have confirmed 
this observation but find an almost 
equal reaction of the human single- 
copy tracer with sea urchin DNA. 
Complementary measurements with 
labeled sea urchin DNA yield similar 
results. Comparable reactions are ob- 
served with DNA of very distant spe- 

cies such as the California mussel (a 
protostome) using either sea urchin or 
human tracer. The hydroxylapatite- 
bound fraction is primarily eluted be- 
tween 50° and 65 °C and is thus either 
very poorly matched or has very short 
regions of sequence homology. There is 
as yet no proof that sequence homol- 
ogy is responsible for the hydroxylapa- 
tite binding, but long incubations are 
required. Various tests indicate that 
the reaction is not due to impurities. 
Measurements are planned with cloned 
fragments which can be sequenced to 
determine the true extent of homology. 


Deininger, P. L., and C. W. Schmidt. 
(1979), J. Mol. Biol. 127: 437-460. 


John W. Roberts, Richard L. Hudspeth, and Roy J. Britten 

In studies of the overall sequence or- 
ganization of human DNA, approxi- 
mately 100 recombinant lambda phage 
were chosen at random from a human 
genome library and examined for the 
presence of repeats. The phage isolates 
were tested both by a screening pro- 
cedure and by the blot hybridization 
method. A substantial fraction of the 
interspersed repetitive sequences of 
the human genome belong to the Alu 
family (Houck et al, 1979). Therefore a 
cloned Alu family repeat (from the 5 ' 
side of the human epsilon globin gene) 
was used as an Alu specific probe. 
Total human genomic DNA, labeled to 
high specific activity by nick transla- 
tion, was used as a probe for all high- 
frequency (greater than 100 copies/ 
genome) repeats, including Alu family 

Of 100 inserts averaging 15 kb in 
length, 86% included one or more Alu 
family repeats. This is consistent with 
a random distribution of Alu repeats 
throughout the genome with an aver- 
age inter-Alu spacing of 7.5 kb; similar 
estimates, obtained by other methods, 
have been published. Studies of restric- 
tion digests showed that at least 38% 
of the recombinants contained one or 
more non-Alu high-frequency repeats. 
10% of the recombinants were found 
entirely to lack high-frequency repeti- 
tive sequences. 

Two recombinant phage have been 
further characterized by restriction map- 
ping and saturation hybridization of 
genomic DNA with isolated restriction 
fragments. One of these recombinants 
includes two high-frequency non-Alu 
repeats interspersed with single-copy or 



low-frequency sequences. The repeats 
appear to differ in frequency and do not 
hybridize with each other. The second 
recombinant consists entirely of single- 
copy and low-frequency repetitive se- 
quences. The work is continuing in an 
attempt to determine the number of 

low-frequency repeats and their distri- 
bution through the genome. 


Houck, C. M., F. P. Rinehart, and C. 
Schmidt. (1979), J. Mol. Biol. 132: 289-306. 



Boning Gao and Eric H. Davidson 

It is known that there are thousands 
of repetitive sequence families in the 
sea urchin genome. RNA transcripts 
from at least some of these families are 
present in both the nucleus and cyto- 
plasm of sea urchin embryo cells. Both 
complements of most repeat sequences 
are represented in the RNA, and the 
steady-state levels of these transcripts 
are tissue-specific. These characteris- 
tics have suggested that repeat se- 
quence transcripts may have some 
function in gene regulation. 

We are studying human repeat se- 
quences to determine whether they 
have the same characteristics. We have 
randomly chosen two clones from a 
human genomic library. The two 
human DNA inserts were digested by 
Hae III endonuclease and the result- 
ing fragments cloned into plasmid 
pBR322. We will select pBR322 sub- 
clones containing repeat sequences and 
plan to look for transcripts of these se- 
quences in human tissue culture cell 


James W. Posakony and Eric H. Davidson 

We are studying the developmental 
pattern of expression of repetitive se- 
quence families in the sea urchin 
genome. It was found previously that 
most repeat families are represented in 
the RNA stored in the mature unfertil- 
ized egg, and that transcripts of both 
strands of each repeat occur at approx- 
imately equal levels. Recently we have 
investigated the nature of these tran- 
scripts in the polyadenylated fraction 
of egg RNA by RNA gel blot experi- 
ments with cloned DNA probes. Each 
repeat family studied was found to be 
represented by multiple large (3-10 
kilobase) transcripts; furthermore, the 
transcripts complementary to the in- 
dividual strands of each repeat are of 
distinct sizes. 

Similar gel blot experiments with 
total polyadenylated RNA from 16-cell 
embryos revealed little difference from 
egg RNA in the pattern of transcripts 
hybridizing to several strand-sepa- 
rated repeat probes. Thus, the prom- 
inent repeat-containing transcripts 
stored in the egg do not appear to 
undergo post-fertilization processing 
events of a kind that would detectably 
alter their size, at least up to five hours 
after fertilization. By contrast, the pat- 
tern of hybridizing transcripts in 
polyadenylated RNA of blastula-stage 
embryos is markedly different from the 
egg RNA pattern for at least one 
strand of all repeats studied, indicating 
a qualitative developmental regulation 
of repeat sequence expression. 



Finally, a combination of electron transcripts in the cytoplasm of em- 
microscopy, RNA gel blots, and solu- bryos. The possibility that these tran- 
tion hybridization experiments has re- scripts appear on polyribosomes for 
vealed the presence of repeat sequence translation is under study. 


David M. Anderson, Margaret E. Chamberlin, and Eric H. Davidson 

It has recently been demonstrated 
that at least half of the mass and most 
of the different single-copy maternal 
mRNA sequences in the sea urchin 
(Strongylocentrotus purpuratus) egg 
are covalently associated with tran- 
scripts of short repetitive sequences 
(Costantini et al, 1980). We have found 
that most (68%) of the poly(A)+ RNA 
transcripts in the full-grown (stage 6) 
Xenopus oocyte also contain at least 
one repetitive element covalently asso- 
ciated with single-copy sequence, and 
at least 50% of these contain two or 
more different repeated elements per 
transcript. In contrast, tadpole cyto- 
plasmic poly(A) + RNA includes few 
(15%) transcripts bearing a repeated 
element. Thus, the covalent associa- 
tion of repeated and single-copy se- 
quences appears to be principally a 
characteristic of the egg poly(A) + 
RNA population. This suggests that 
the repetitive-element-containing tran- 
scripts represent unprocessed mes- 
senger RNA precursors. We have frac- 
tionated sea urchin egg poly(A) + RNA 
into repeat-containing and non-repeat- 
containing fractions and compared 
their translatability in vitro. Incorpor- 

ation of 35 S-methionine is about five 
times greater per mg of input RNA in 
the non-repeat-containing fraction. 

We have also determined that the 
newly synthesized poly(A) + RNA tran- 
scripts present in the nucleus and cyto- 
plasm of stage 6 Xenopus oocytes are 
indistinguishable in structure and se- 
quence content from those transcripts 
that have accumulated during or just 
prior to the maximal lampbrush stage 
(stage 3) of oogenesis. This demon- 
strates that the stage 6 oocyte nucleus 
actively transcribes and exports to the 
cytoplasm transcripts identical in 
structure and sequence content to 
those accumulated earlier in oogenesis. 
Therefore, the maternal poly(A) + RNA 
population in the Xenopus laevis 
oocyte cytoplasm must be at steady 
state with regard to synthesis and 
turnover, as suggested by Dolecki and 
Smith (1979). 


Costantini, F. D., R. J. Britten, and E. H. 

Davidson. (1980), Nature 287: 111-117. 
Dolecki, F. J., and L. D. Smith. (1979), Devel 

Biol. 69: 217-236. 


Howard T. Jacobs, Terry L. Thomas, and Eric H. Davidson 

Using conventional blotting technol- 
ogy, we have identified a number of 
cloned single-copy genes represented in 
the maternal RNA of the sea urchin 
egg predominantly in the form of tran- 

scripts much larger than the corre- 
sponding native mRNA. These include 
species found at high prevalence on the 
polysomes during early development. 
Our studies are directed toward es- 



tablishing whether these represent 
"prespliced" mRNA precursors and 
whether they are mobilized by RNA 
processing after fertilization. 

To elucidate the structure of mater- 
nal transcripts, we are employing a 
combination of blot hybridization and 
nucleotide sequencing, using frag- 
ments of the relevant genes derived 
from genomic clones and the appro- 
priate set of clones obtained from a 
randomly primed cDNA library gener- 
ated from maternal RNA. The method 
of nucleotide sequencing we have 
adopted is that of primer extension, us- 
ing a set of overlapping short frag- 
ments cloned in the single-stranded 
phage Ml 3. 

To determine whether maternal 
RNA is subsequently processed to gen- 
erate functional messages, we are ex- 
ploiting the fact that during early de- 
velopment, several highly prevalent 
mRNAs are released to the cytoplasm 
at a very low rate compared with their 
steady-state concentrations. Thus, by 
measuring the rate of decay of their 
coding and intervening sequences dur- 
ing this period wherein de novo tran- 
scription does not contribute signifi- 
cantly to the mass of RNA, we should 
be able to establish the fate of the 
prespliced maternal transcripts. 


Barbara R. Hough-Evans, Henry Niman, * and Eric H. Davidson 

The vitelline layer (VL) of the sea 
urchin egg lies over the plasma mem- 
brane and consists of a number of pro- 
teins and glycoproteins including a 
sperm attachment site. At fertilization 
the sperm penetrates the vitelline layer 
before fusing with the egg. Shortly 
thereafter the vitelline layer is elevated 
from the egg surface and becomes the 
fertilization membrane, within which 

*Scripps Clinic and Research Foundation, La 
Jolla, California. 

the embryo develops for the next 18 
hours until it hatches. We have in- 
jected vitelline layers prepared from 
unfertilized eggs into rabbits, and ob- 
tained a mixture of anti-VL antibodies. 
Monoclonal antibodies are also being 
prepared, against individual VL pro- 
teins. We plan to characterize the in- 
dividual antibodies and use them to 
identify cloned VL protein genes in a 
recombinant DNA library. This library 
is in the process of being prepared from 
cDNA homologous to the cytoplasmic 
poly(A) RNA of sea urchin ovaries. 


Ji-hou Xin, Bruce P. Brandhorst, and Eric H. Davidson 

This study was undertaken to deter- 
mine whether genes present in cyto- 
plasmic poly(A) RNA of sea urchin em- 
bryos at moderate and high prevalence 
are also represented in messenger 
RNA of coelomocytes, the cells of the 

adult coelomic fluid. The measure- 
ments were carried out by the colony 
hybridization method, using 455 clones 
of pluteus stage embryo cDNAs, and 
labeled cDNA from coelomocyte 
poly (A) RNA. It was found that only 



5.5% of the pluteus clones are present 
in coelomocyte mRNA at levels that 
can be detected. That is, 94.5% of these 
genes are represented by less than one 
transcript per coelomocyte. Most of 
the clones represented in coelomocyte 
RNA come from the high-prevalence 
embryo transcript group, and a few are 
from the group of moderately pre- 

valent transcripts in the pluteus. We 
conclude that they may represent sea 
urchin genes that are expressed ubiqui- 
tously. It is clear, however, that the 
majority of genes giving rise to moder- 
ately prevalent embryo mRNAs must 
be regulated during the life cycle of the 
sea urchin. 


Constantin N. Flytzanis and Eric H. Davidson 


For most poly(A) + RNA sequences 
of the sea urchin, the prevalence levels 
determined during oogenesis are main- 
tained throughout early development, 
while a minority of sequences display 
sharp, stage-specific changes. To study 
such changes, cDNA libraries made 
from different embryonic stages have 
been screened with labeled cDNAs 
transcribed from polysomal poly(A) + 
RNAs of 16-cell, blastula, gastrula, 
and pluteus stage embryos. The con- 
centration of the poly(A) + RNAs com- 
plementary to our clones in these em- 
bryonic stages has been calculated, 
and a set of clones that show dramatic 
changes has now been selected for fur- 
ther investigation. We have again 
screened this set of clones with cDNAs 
transcribed from poly(A) + RNAs of 
early embryo and adult tissues. Three 
cDNA clones have been selected that 
are complementary to RNAs appear- 
ing in the cytoplasm of the sea urchin 
embryo but not detectable with this 
method in the mature egg and the 

adult tissues tested. Using these clones 
as probes for genomic blot analysis, we 
have shown that the respective RNAs 
are transcribed from single-copy se- 
quences of the sea urchin genome. Us- 
ing RNA blot analysis, we have ana- 
lyzed the nature of the transcripts 
throughout development. For each of 
the three clones, only one transcript 
could be identified in the cytoplasmic 
RNA of the blastula, gastrula, and 
pluteus stages. Though these tran- 
scripts were absent in the total RNA of 
the adult tissues, other, longer tran- 
scripts of lower concentration could be 
detected. The possibility that these 
transcripts represent nuclear nonpro- 
cessed RNAs is currently under inves- 
tigation. We hope that the study of 
these transcripts in the nucleus of the 
embryo and the adult tissues will eluci- 
date the mechanisms on the transcrip- 
tional or post-transcriptional level that 
control the expression of these se- 
quences specifically in the embryonic 




Bruce P. Brandhorst, Constantin N. Flytzanis, and Eric H. Davidson 

The population of rare polysomal 
mRNAs changes during embryonic de- 
velopment of S. purpuratus. That there 
are also changes in the population of 
more prevalent mRNAs is indicated by 
changes in the patterns of protein syn- 
thesis detected by two-dimensional gel 
electrophoresis (Brandhorst, 1976). A 
large fraction of these changes occur 
shortly after hatching and are accom- 
panied by similar changes in the popu- 
lation of cytoplasmic RNA translat- 
able in cell-free translation systems (A. 
Bedard and B. Brandhorst, unpub- 
lished observations). It is thus possible 
that there is a coordinate induction of 
the expression of a group of genes pre- 
ceding mesenchyme formation and the 
beginning of gastrulation. Our present 
research is exploring this possibility 
further. We are using colony hybridiza- 
tion to screen several cDNA libraries 
with cDNA transcribed from cyto- 
plasmic polyadenylated RNA of var- 
ious embryonic stages. The abun- 
dances of the RNAs coded for by the 
screened cloned DNAs can be esti- 

mated from standard curves, according 
to Lasky et al. (1980). They conclude 
that most poly(A) + RNA transcripts 
prevalent in the egg (10 4 copies or more 
per egg) remain prevalent throughout 
embryonic development, though quan- 
titative changes are observed. A com- 
mon pattern is the decline in preva- 
lence through early embryonic develop- 
ment followed by a recovery after 
gastrulation; this implies that mater- 
nal RNAs are decaying and eventually 
are replaced by newly transcribed 
RNA in older embryos. Few new preva- 
lent poly(A) + RNAs appear during em- 
bryonic development. Prevalent RNAs 
detected in the cytoplasm are also 
prevalent in polysomes at the 16-cell 
and gastrula stages. 


Brandhorst, B. P. (1976), Devel. Biol. 52: 

Lasky, L. A., Z. Lev, J.-H. Xin, R. J. Britten, 

and E. H. Davidson. (1980), Proc. Nat. Acad. 

Sci. USA 77: 5317-5321. 


Jay W. Ellison, Carlos V. Cabrera, James G. Moore, * and Eric H. Davidson 

We have investigated the metabo- 
lism of individual cytoplasmic poly(A) 
RNA species in sea urchin embryos in 
order to learn about the factors control- 
ling their prevalence. The sequences in 
our study differ broadly in abundance 
from 0.5 to 1500 molecules per cell. Our 
approach has been to add 3 H- 
guanosine to cultures of embryos at 
the blastula-gastrula stage of develop- 
ment and to prepare cytoplasmic 

*Xerox Corp., Pasadena, California. 

poly(A) RNA from embryos harvested 
at various times after addition of the 
isotope. Culture samples were also re- 
moved for the determination of the 
GTP pool specific activity. The 
3 H-RNA was hybridized to cloned 
cDNA sequences. The results of such 
experiments, combined with the GTP 
pool measurements, allowed us to de- 
termine absolute rates of cytoplasmic 
entry and decay for several individual 
transcripts. These studies showed that 
rare RNA species turn over rapidly, 



and their prevalence depends on the 
flow rate of the new transcripts into 
the cytoplasm. On the other hand, 
more abundant transcripts are very 
stable, this being the main factor in 
maintaining their prevalence. The 
prevalent RNAs fall into two groups 
with respect to their rates of appear- 

ance in the cytoplasm. Some are trans- 
ported to the cytoplasm at rates high 
enough to account for all the tran- 
scripts present as newly synthesized. 
For others the "de novo" appearance 
only accounts for 1-10% of the total 
cytoplasmic transcripts— thus the rest 
persist from the maternal RNA. 


Carlos V. Cabrera, James J. Lee, James G. Moore, * and Eric H. Davidson 

Following the approach of Ellison et 
al (Year Book 79, p. 222), we are mea- 
suring transcription rates for a variety 
of RNA sequences which differ broadly 
in their abundance in the cytoplasm. 
Our purpose is to gain information 
about the contribution of post-tran- 
scriptional processes (occurring in the 

*Xerox Corp., Pasadena, California. 

nucleus) to the RNA prevalence in the 

Our present experiments are directed 
toward uncovering whether the kinetic 
parameters (rate of synthesis and 
decay) and/or the efficiency of trans- 
port to the cytoplasm change through- 
out development. Results from this 
work should also provide insights into 
the control of prevalence classes in the 
cytoplasmic RNA population. 



Terry L. Thomas and Eric H. Davidson 

A 14-kilobase (kb) fragment of sea ur- 
chin DNA cloned in bacteriophage 
lambda was chosen for detailed investi- 
gation by screening a library of cloned 
genomic DNA. This region of the 
genome includes the transcription unit 
of a previously isolated cloned DNA se- 
quence, Sp88, and encodes four poly- 
adenylated RNA species (Lee et al, 
1980; Lev et al, 1980). One of these 
species, 10 kb in length, is present in 
mature oocyte RNA. This transcript is 
co-linear with transcripts found in nu- 
clei of blastula- and gastrula-stage em- 
bryos, and it has an interspersed se- 

quence organization (i.e., both repeated 
and single-copy DNA sequences are 
represented in the RNA molecule). The 
remaining Sp88 transcripts are 4, 3, 
and 1.8 kb in length and are present in 
maternal RNA and on polysomes of 
16-cell embryos. The 10-kb transcript 
overlaps the three polysomal species at 
their 3 ' termini. In addition, this tran- 
script contains at least one intervening 
sequence. Thus, the 10-kb maternal 
transcript of gene Sp88 resembles a 
nonprocessed precursor to the three 
smaller mRNAs. From this and other 
observations, we conclude that nonpro- 



cessed transcripts are present in sea ur- 
chin maternal RNA. We do not know 
the fate or function of these nonpro- 
cessed maternal RNAs. However, fur- 
ther investigation of this novel exam- 
ple of gene expression may provide in- 
sights into questions of control of gene 
expression in early embryogenesis. 


Lee, A. S., T. L. Thomas, Z., Lev, R. J. Britten, 

and E. H. Davidson. (1980), Proc. Nat. Acad. 

Sci. USA 77: 3259-3263. 
Lev, Z., T. L. Thomas, A. S. Lee, R. C. Angerer, 

R. J. Britten, and E. H. Davidson. (1980), De- 

vel. Biol. 76: 322-340. 


David H. Price and Eric H. Davidson 

Gene expression in the sea urchin is 
controlled to a certain extent at the 
level of transcription. An in vitro tran- 
scription system using isolated nuclei 
from sea urchin embryos was charac- 
terized. Depending upon the method of 
isolation, the nuclei exhibited tran- 
scription rates of about 10-20% of that 
calculated from in vivo measurements. 
Approximately 90% of the total tran- 
scriptional activity is inhibited by low 
levels of a-amanitin, indicating that 
polymerase II is responsible for the 
vast majority of the RNA being made 
in early embryogenesis. The tran- 
scripts synthesized in vitro are hetero- 
geneous in size, ranging from 100 to 
more than 20,000 bases with weight 
average around 8,000 bases. Transcrip- 
tion in the presence of low levels of 
a-amanitin results in the production of 
only one high-molecular-weight tran- 
script of about 8,000 bases, which 

is presumably the ribosomal RNA 
(rRNA) precursor. No processed rRNA 
has been detected in vitro. 

The in vitro transcription rates for 
several specific genes were determined 
by hybridization of the in vitro- 
synthesized RNA to filters containing 
cloned DNA complementary to specific 
sea urchin transcripts. While genes 
that are normally transcribed at rela- 
tively low rates in vivo had correspond- 
ingly low in vitro rates, rapidly tran- 
scribed genes had considerably lower 
rates in vitro than in vivo. This dispar- 
ity could be due to the lack of reinitia- 
tion in vitro, which would affect the 
transcription of genes loaded with 
many polymerase molecules more than 
genes containing on the average only 
one or several polymerase molecules. It 
is hoped that this system can be used 
to examine the processing of specific 
transcripts in vitro. 


Richard H. Scheller, Linda B. McAllister, * William R. Crain, Jr.,f David S. Durica,f 
James W. Posakony, Terry L. Thomas, and Eric H. Davidson 

A set of at least eleven actin genes 
has been isolated from genomic recom- 
binant DNA libraries of the sea urchin 
Strongylocentrotus purpuratus. Most 

♦Undergraduate, California Institute of Tech- 

•fThe Worcester Foundation for Experimental 
Biology, Shrewsbury, Massachusetts. 

of the isolates derive from a library rep- 
resenting the genome of a single an- 
imal. There are at least five distinct 
types of sea urchin actin gene, some of 
which are represented by multiple cop- 
ies in the genome. The actin gene types 
are distinguished by nonhomologous 
flanking sequences and intervening se- 
quences, though the protein coding se- 



quences appear in most cases to be 
quite similar. Eight of the eleven genes 
isolated have been recovered in X re- 
combinants containing two actin genes, 
linked at 5-9 kilobase (kb) distances. 
Restriction map overlaps suggest that 
the genome contains an array of at 
least three of these genes spaced over 
about 30 kb of DNA. In the linkage 
patterns observed, actin genes of 
diverse types are linked to each other. 
In early embryos, actin mRNA tran- 
scripts of 1.8 and 2.2 kb are found, and 
the longer of these transcripts is more 
prevalent in the maternal RNA of the 

egg. From RNA gel blot experiments, 
we conclude that the two transcripts 
derive from different actin gene types. 
Different repetitive sequences are 
located either side of most of the actin 
genes, and in most observed cases the 
repeat sequences adjacent to actin 
genes of a given type are similar. The 
repeat sequences flanking the actin 
genes belong to families that are tran- 
scribed, but those repeats in the neigh- 
borhood of the actin genes that have 
been investigated are not themselves 
represented in the stable RNAs of eggs 
or early embryos. 


Terry L. Thomas, Margaret P. Price, 

Linda B. McAllister, * Juanito S. Villanueva, * and 

Eric H. Davidson 

There are five distinct types of actin 
genes in the sea urchin genome 
(Scheller et al, 1981; Year Book 79, p. 
216.) Multiple overlapping lambda re- 
combinants suggest that at least three 
actin genes are linked and spaced over 
approximately 30 kb of DNA. The 
linked actin genes are heterotypic. 
RNA gel blots using actin genes as 
probe show two transcripts, 2.2 and 1.8 
kb in length, in eggs and early em- 
bryos. The mass ratio of these two 
transcripts changes during early devel- 
opment. Although there are only two 
size classes of actin transcript, these 
probably include several distinct actin 
mRNAs. The actin coding sequences of 
the different genes are quite similar, 
but the 3 ' DNA sequences encoding 
untranslated regions of actin mRNAs 
differ substantially. We are using these 

"■Undergraduate, California Institute of Tech- 

3 ' sequence differences to identify the 
actin gene types with respect to the dif- 
ferent actin mRNAs. At present we 
have identified at least three genes 
that give rise to different actin mRNAs 
in eggs and early embryos. It is prob- 
able that additional actin genes encod- 
ing mRNAs that are tissue- and/or 
stage-specific will be identified as new 
3' specific probes are used. We will 
confirm our gene assignments by 
direct sequencing of 3' untranslated 
regions of cDNA clones isolated with 
the various 3 ' specific probes. In addi- 
tion, we are screening bacteriophage 
lambda and cosmid libraries with actin 
probes to determine if there is longer 
range linkage of actin genes in the sea 
urchin genome. 


Scheller, R. H., L. B. McAllister, W. R. Crain, 
D. S. Durica, J. W. Posakony, T. L. Thomas, 
R. J. Britten, and E. H. Davidson. (1981), Mo- 
lecular & Cellular Biology, in press. 




Lajos Piko 

One of the earliest events of mouse 
embryo differentiation is the produc- 
tion of large numbers of intracisternal 
A-type particles— small virus-like struc- 
tures closely associated with the endo- 
plasmic reticulum membrane. The parti- 
cles are absent in the one-cell fertilized 
egg but appear in large clusters, about 
1-2 X 10 5 particles per embryo, at the 
two-cell stage; this number is reduced 
to about 1000 per embryo by the early 
blastocyst stage (32 cells). This pattern 
of expression has been observed in 
about a dozen laboratory strains and in 
feral mice (Piko, 1975). 

On the basis of morphology and im- 
munologic cross-reactivity, the A 
particles in the embryos appear to be 
related to similar particles abundant in 
mouse myelomas and some other 
tumors. To pursue this relationship 
further, we have produced 32 P-labeled 
cDNA to A-particle RNA from my- 
elomas and used it to screen the mouse 
genomic library (in collaboration with 
Mark Davis of Lee Hood's group). 
About 1% of the clones reacted, in 
agreement with the reported high reit- 
eration of A-particle gene (about 1000 

copies per haploid mouse genome; 
Leuders and Kuff, 1977). Ten clones 
containing A-particle genes have been 
analyzed by restriction enzyme cleav- 
age and heteroduplex mapping. All 
genes seem to have similar terminal re- 
peats but fall into three general pat- 
terns as to size and sequence organiza- 
tion (in substantial agreement with the 
findings of Ono et al, 1980): typical 
large genes, about 7 kb in length (three 
clones); typical small genes, about 4.5 
kb, with portions of the large gene 
missing (three clones); and atypical 
genes, 2.5-5.0 kb, with extensive sub- 
stitutions and deletions. We plan to 
study the sequence relationship be- 
tween these clones and A-particle RNA 
from mouse embryos. 


Leuders, K. K., and E. L. Kuff. (1977), Cell 12: 

Ono, M., M. Cole, A. T. White, and R. C. C. 
Huang. (1980), Cell 21: 465-473. 

Piko, L. (1975), in The Early Development of 
Mammals, M. Balls and A. E. Wild, eds., pp. 
167-187, Cambridge University Press, Cam- 


Anderson, D. M., R. H. Scheller, J. W. 
Posakony, L. B. McAllister, S. W. Trabert, 
C. Beall, R. J. Britten, and E. H. Davidson. 
(1981), Repetitive sequences of the sea urchin 
genome. Distribution of members of specific 
repetitive families. J. Mol. Biol. 145: 5-28. 

Anderson, D. M., J. Richter, L. D. Smith, M. E. 
Chamberlin, R. J. Britten, and E. H. 
Davidson. (1981), Sequence structure of the 
poly(A) + RNA synthesized and accumulated 
during oogenesis in the amphibian Xenopus 
laevis. Manuscript in preparation. 

Costantini, F. D., R. J. Britten, and E. H. David- 
son. (1980), Message sequences and short re- 

petitive sequences are interspersed in sea ur- 
chin egg poly (A) + RNAs. Nature 287: 111-117. 

Ellison, J. W., C. V. Cabrera, J. G. Moore, 
R. J. Britten, and E. H. Davidson. (1981) 
Control of mRNA prevalences during early 
sea urchin embryogenesis: labeling kinetics 
of specific mRNA sequences. Manuscript 
in preparation. 

Ernst, S. G., B. R. Hough-Evans, R. J. Britten, 
and E. H. Davidson. (1980), Limited 
complexity of the RNA in micromeres of 
sixteen-cell sea urchin embryos. Deuel. 
Biol. 79: 119-127. 



Grula, J. W., T. J. Hall, J. A. Hunt, 
T. D. Giugni, E. H. Davidson, and R. J. 
Britten. (1981), Sea urchin DNA sequence 
polymorphism and reduced interspecies 
differences of the less polymorphic DNA 
sequences. Manuscript in preparation. 

Hall, T. J., J. W. Grula, E. H. Davidson, 
and R. J. Britten. (1980), Evolution of sea ur- 
chin non-repetitive DNA. J. Mol. Evol 16: 

Hough-Evans, B. R., M. Jacobs-Lorena, 
M. R. Cummings, R. J. Britten, and E. H. 
Davidson. (1980), Complexity of RNA in 
eggs of Drosophila melanogaster and Musca 
domestica. Genetics 95: 81-94. 

Jacobs, H. T., T. L. Thomas, J. W. Posakony 
B. R. Hough-Evans, R. J. Britten, and E. H. 
Davidson. (1981), Mechanisms of eukaryotic 
gene regulation, in Perspectives in Differentia- 
tion and Hypertrophy, Elsevier Press, Am- 
sterdam, in press. 

Lasky, L. A., Z. Lev, J.-H. Xin, R. J. 
Britten, and E. H. Davidson. (1980), Messen- 
ger RNA prevalence in sea urchin embryos 
measured with cloned cDNAs. Proc. Nat. 
Acad. Sci. USA 77: 5317-5321. 

Lasky, L. A., Z. Lev, T. L. Thomas, 
J.-H. Xin, A. S. Lee, R. J. Britten, and 
E. H. Davidson. (1981), The expression of 
abundant and rare mRNA sequences during 
sea urchin development, in Progress in 
Developmental Biology, H. W. Sauer, ed., 
Elsevier Press, Amsterdam, in press. 

Leahy, P. S., B. R. Hough-Evans, R. J. 
Britten, and E. H. Davidson. (1981), Syn- 
chrony of oogenesis in laboratory-maintained 

and wild populations of the purple sea urchin 
(Strongylocentrotus purpuratus). J. Exp. Zool. 
215: 7-22. 

Lee, A. S., T. L. Thomas, Z. Lev, R. J. 
Britten, and E. H. Davidson. (1980), Four 
sizes of transcript produced by a single sea 
urchin gene expressed in early embryos. Proc. 
Nat. Acad. Sci. USA 77: 3259-3263. 

Posakony, J. W., R. H. Scheller, D. M. 
Anderson, R. J. Britten, and E. H. Davidson. 
(1981), Repetitive sequences of the sea urchin 
genome. Nucleotide sequences of cloned 
repeat elements. J. Mol. Biol., in press. 

Scheller, R. H., D. M. Anderson, J. W. 
Posakony, L. B. McAllister, R. J. Britten, 
and E. H. Davidson. (1981), Repetitive se- 
quences of the sea urchin genome. Subfamily 
structure and evolutionary conservation. J. 
Mol.Biol, in press. 

Scheller, R. H., L. B. McAllister, W. R. 
Crain, D. S. Durica, J. W. Posakony, T. L. 
Thomas, R. J. Britten, and E. H. Davidson. 
(1981), Organization and expression of 
multiple actin genes in the sea urchin. 
Molecular & Cellular Biology, in press. 

Thomas, T. L., J. W. Posakony, D. M. 
Anderson, F. D. Costantini, R. J. Britten, 
and E. H. Davidson. (1981), Novel molecular 
structure of maternal RNA. Chromosoma,ix\ 

Xin, J.-H., B. P. Brandhorst, R. J. Britten, 
and E. H. Davidson. (1981), Cloned embryo 
mRNAs not detectably expressed in adult sea 
urchin coelomocytes. Submitted for publica- 


Professor: Eric H. Davidson 

Distinguished Carnegie Senior Research 
Associate: Roy J. Britten 

Senior Research Associate: Barbara R. 

Visiting Associates: Bruce P. Brandhorst, 
Lajos Piko,* Kwei-fang Ts'ui, Ji-hou Xin 

*Veterans Administration Medical Center, 
Sepulveda, California. 

Senior Research Fellows: David M. Ander- 
son, John W. Roberts, Terry L. Thomas 

Gosney Research Fellow: Ze'ev Lev 

Research Fellows: Carlos V. Cabrera, 
Constantin N. Flytzanis, John W. Grula, 
Terrence J. Hall, Howard T. Jacobs, 
Steven A. Johnson, Laurence A. Lasky, 
David H. Price, Richard H. Scheller 

Graduate Students: Jay W. Ellison, 
Boning Gao, James L. Lee, James W. 



Research Staff: Maria Alonso, Alison 
E. Blake, Margaret E. Chamberlin, 
Geoffrey J. Graham, Patrick S. Leahy, 
Jara Lewin, Margaret Price, Jane Rigg, 
Nelson Smith 

Laboratory Staff: Carlzen Balagot, Fargo 
Balliett, Joseph A. Garcia, Jr., Richard 
L. Hudspeth, Terrence Giugni, Linda B. 
McAllister, Juanito Villanueva 


The work described in the preceding 
research reports has been supported by: 

Biomedical Research Support Grant (NIH) 
California Foundation for Biochemical 

Chinese Embassy (People's Republic of 

Norman W. Church Foundation 
Charles B. Corser Fund for Biological 


Deutsches Krebsforschungszentrum 

Fogarty International Center (NIH) 

E. S. Gosney Fund 

McGill University, Canada 

National Institutes of Health, USPHS 

National Science Foundation 

Science Research Council Fellowship, 

Veterans Administration 

Geophysical Laboratory 

Washington, District of Columbia 

Hatten S. Yoder, Jr. 


Commentary and Overview of 
the Director 251 

Experiments Bearing on the Earth's Lower 

Mantle and Core 267 

High-pressure experiments on FeS with 
bearing on the composition of the 
earth's core (Mao, Zou, and Bell) . . . 267 

Application of the solid-helium pressure 
medium in a study of the at-e Fe 
transition under hydrostatic pressure 
(Zou, Bell, and Mao) 272 

Irreversible unit-cell volume changes of 
wiistite single crystals quenched from 
high pressure (Hazen, Mao, Finger, 
and Bell) 274 

Systematic variation of bulk modulus of 
wiistite with stoichiometry (Hazen) . . 277 

Equations of state for thorium metal, 
U0 2 , and a high-pressure phase of 
U0 2 to 650 kbar (Benjamin, Zou, Mao, 
and Bell) 280 

High-pressure phase transformations in 
selenium and tellurium (Mao, Zou, and 
Bell) 283 

Phase transition in GaAs at high 
hydrostatic pressures (Bell and Mao) . 284 

Elasticity data on hydrogen and 
deuterium at high pressure (5-200 
kbar) (Mao, Brody, Shimizu, and Bell) 286 

The HD reaction at high pressure (Bell, 
Mao, and Sharma) 294 

Isothermal equations of state for neon 
and argon (Zou, Finger, Hazen, Bell, 
and Mao) 295 

Liquid and Glass Structures 300 

Structural similarity between melts and 
glass relevant to penological 
processes (Seifert, Mysen, and 
Virgo) 300 

Quantitative determination of 
proportions of anionic units in silicate 
melts (Seifert, Mysen, and Virgo) . . 301 

Distribution of aluminum between 
structural units in silicate melts 
(Mysen, Virgo, and Kushiro) 302 

Structure and properties of 
aluminosilicate melts with three- 
dimensional network structure (Seifert, 
Mysen, and Virgo) 305 

Relationship between the oxidation state 
of iron and the structure of silicate 
melts (Virgo, Mysen, and Seifert) . 308 

Ferric iron as a network former and as a 
network modifier in melts relevant to 
penological processes (Mysen, Virgo, 
and Seifert) 311 

The oxidation state of europium in albite 
and alkali-earth silicate glasses (Virgo, 
Seifert, and Mysen) 313 

Igneous Petrology 317 

A liquid-density controlled model for 
chromitite formation in the Muskox 
Intrusion (Irvine) 317 

Some additional observations on magnetite 
in calc-alkaline volcanic rocks (Osborn 
andBoctor) 324 

Quenched rocks of mantle origin from the 
Mzongwana kimberlite dike, Transkei, 
southern Africa (Boyd, Nixon, and 
Boctor) 328 

The pressure dependence of the viscosity 
of some basic melts (Scarfe) 336 

Change in viscosity with pressure of melts 
in the system CaO-Al 2 3 -Si0 2 
(Kushiro) 339 

Preliminary experimental results in the 
system Li AlSi0 4 -Si0 2 -H 2 O (London) . 341 

Characterizing associations between 
proportions by means of remaining 
space variables (Chayes) 345 

Element Partitioning 347 

Rare earth element partitioning between 
minerals and (C0 2 + H 2 0) vapor as a 
function of pressure, temperature, and 
vapor composition (Mysen) 347 

Phase relations in the system 
Na 2 0-Al 2 3 -Si0 2 -H 2 0-HF at 15 kbar 
(Danckwerth) 350 

Partitioning of nickel and cobalt between 
pyrrhotite, pyrite, and magnetite 
(Boctor) 352 

Partitioning of nickel between olivine and 
iron monosulfide melts (Boctor) . . . 356 

Fluid Speciation and Dynamics 359 

Mineral-solution equilibria in the system 
MnO-Si0 2 -HCl-H 2 (Boctor and 

Frantz) 359 

Ionization constants of calcium chloride 
and magnesium chloride in aqueous 
solutions at temperatures to 600 °C and 
pressures to 4000 bars (Frantz and 

Marshall) 362 

Oxygen isotope equilibration during 
regional metamorphism (Rumble and 
Hoering) 367 

Comparative Crystal Chemistry 370 

Refinement of the crystal structure of an 
iron-rich kornerupine (Finger and 
Hazen) 370 

Crystal structure of diopside at high 
temperature and pressure (Hazen and 
Finger) 373 

Cell parameters of orthoenstatite at high 
temperature and pressure (Ralph, 
Hazen, and Finger) 376 

The crystal structure of bannermanite 
(Hughes, Finger, and Hazen) 379 

Crystal structure of pyroxmangite at high 
temperature (Pinckney, Finger, Hazen, 
and Burnham) 380 

Biogeochemistry 385 

Isotopic composition of hydrogen and 
carbon in mat-forming, thermophilic 
algae and bacteria (Estep) 385 

Aminostratigraphy of the South 
Carolina Coastal Plain by monospecific 
fossil analyses (Corrado and Hare) . . 387 

Monomethyl, acyclic hydrocarbons in 
petroleum and rock extracts (Hoering) 389 

A new chromatographic procedure for the 
resolution of mixtures of acidic, neutral, 
and basic amino acid enantiomers 
(Engel and Hare) 394 

A procedure for the isolation and 
resolution of d- and L-tryptophan in 
peptides and proteins (Engel) .... 397 

Determination of the extent of 
racemization of basic amino acids in 
fossil shells with high-performance 
liquid chromatography (Hare and 
Engel) 398 

Resolution of d- and L-enantiomers of 
arginine, histidine, lysine, and 
tryptophan in heat-alkali-treated 
a-melanocyte- stimulating hormone 
(Engel and Hare) 400 

New Techniques and Methods 404 

Degrees of hydrostaticity in He, Ne, and 
Ar pressure-transmitting media (Bell 
and Mao) 404 

High-pressure, high-temperature optical 
cell for Raman spectroscopy (Frantz) . 406 

Thermocouple controller for high- 
temperature furnace (Hadidiacos) 408 

Evaluation of a method for measuring 
the isotopic composition of oxygen in 
organic matter (Hoering and Estep) . . 410 

Selective adsorption of organic molecules 
by silicalite (Freeman) 414 

Staff Activities 416 

Conference on the Geochemistry of 

Organic Matter in Ore Deposits . . . 416 

Seminar Series 417 

Penologists' Club 418 

Washington Crystal Colloquium . . . .418 

Lectures and Symposia 418 

Field Studies 421 

Bibliography 422 

Personnel 424 


Seventy-five years ago ground was 
broken for the Geophysical Laboratory. 
The exciting concept that was planted 
and nurtured— to undertake funda- 
mental research on the chemistry and 
physics of the earth— has flourished 
ever since. Underlying the concept is 
the freedom of the individual investiga- 
tor to evolve and grow, to seek new so- 
lutions to major scientific problems, to 
create and innovate, to pioneer and in- 
vent. Although the research, and the 
science itself, is continuously chang- 
ing, the concept requires constancy in 
excellence, creativity, accuracy, and ac- 
countability. Such ideals obtain only 
with a staff that is dedicated intellectu- 
ally, highly motivated, and industri- 
ous; each member must be an inde- 
pendent yet cooperative leader, and all 
must be backed by a trusting and sup- 
portive administration. Nevertheless, 
the work of even the most productive 
staff must be constantly enhanced by 
the enthusiasm of exceptional young 
scholars, by the perspectives of estab- 
lished visiting scientists, by the wisdom 
of seasoned retirees. Inherent in the 
concept of the Laboratory is the flexi- 
bility of the investigators to change di- 
rection and pursue the most illuminat- 
ing and productive areas of research. 

It is indeed remarkable that this in- 
dependent, nonprofit, scientific organi- 
zation, remaining essentially the same 
size after 75 years, has retained all the 
vigor and vitality of its inception. Evo- 
lution of the Geophysical Laboratory's 
scientific programs takes place not to 
meet present-day societal problems, 
but to prepare for tomorrow's needs. 
The principal purpose is to provide the 
groundwork of principles that will help 
solve the myriad of undefined prob- 
lems of the future. 

Because it is not possible to predict 
where opportunities for research will 
lie in the future, one can seek only to 
pursue those sparks of creativity, 

genius, intuition, perception, and inno- 
vation that the staff provide within the 
constraint of available financial re- 
sources. At present, the rate-determin- 
ing step is financial resources— not 
ideas. With that constraint one must 
select carefully those projects that in 
their broad application by others will, 
in Andrew Carnegie's words, most "ben- 
efit mankind." Whatever the financial 
resources, a single organization cannot 
pursue all potentially rewarding in- 
vestigations, and it is essential, there- 
fore, to focus resources and talents on 
projects from which principles of last- 
ing value are likely to emerge. 

The research topics described below 
reflect the judgments by each staff 
member as to the wisest, most benefi- 
cial, and resourceful use of his or her 
talents, individually and in concert. 
The continuity and interrelationships 
can be perceived, however, only by im- 
mersion in the exciting environment 
generated by the staff. The following 
brief notes give a limited view of this 
unique organization as it celebrates its 
75th year of service to basic research 
and education. 

Because of space limitations not all 
studies are described in detail in the 
main body of the Report. Articles de- 
scribed only in this introduction are 
marked with an asterisk. Results that 
will appear in appropriate journals be- 
fore publication of this Report, how- 
ever, have not been included. Titles of 
published papers appear in the bibliog- 

Experiments Bearing on the Earth 's 
Lower Mantle and Core 

Conditions equivalent to the core- 
mantle boundary of the earth can now 
be studied in the laboratory. When ma- 
terials of the upper mantle, as deduced 
in part from the phases in nodules from 
kimberlites, are compressed, they un- 




dergo further reactions and phase 
changes. The minerals produced con- 
sist of a few simple structures, closely 
packed and of extensive solid solu- 
tions. The characterization of these 
structures and their compositions, the 
correlation of their properties with the 
properties of the core and mantle as de- 
termined by geophysical techniques, 
and an understanding of their bearing 
on earth processes are the principal 
goals of the experimenter. Materials 
examined this year include metals, ox- 
ides, sulfides, and arsenides. A tech- 
nique has been demonstrated that 
yields the elastic properties of crystals 
at very high pressure. In addition, a 
critical test of the finite-strain equa- 
tion of state of a crystalline material 
has been made, over an extensive ex- 
perimental range. These accomplish- 
ments result in many new opportunities 
for geophysicists and physicists to 
study materials under the extreme con- 
ditions of the lower mantle and core. 

Mao, Zou, and Bell began an experi- 
mental study of metallic iron and iron 
sulfide (FeS) at very high pressures, 
because those materials are presumed 
to be among the principal minerals of 
the earth's core. They observed the a-e 
transition in iron and the sluggish be- 
havior of the reaction at hydrostatic 
pressures up to approximately 160 
kbar. In FeS (troilite), two transitions 
were noted (at 48 and 130 kbar) that 
may shift the calculated FeS eutectic 
point so often used as a constraint in 
geophysical models of the core. The in- 
vestigators determined the equations 
of state of the high-pressure phase of 
FeS to 600 kbar and calculated values 
for the sulfur content of the core to be 
approximately 7-9 wt %. Combination 
of the FeS data with shock-wave data 
on FeS 2 and Fe 09 S and with their 
previous data on metallic Fe (the high- 
pressure e phase) revealed a trend of 
convergence of the structures to a 
dense, close-packed configuration of 
the atoms. At pressures above 600 
kbar the density is, to a first approxi- 

mation, sensitive to composition alone. 
Comparison of the specific volumes 
yields a linear, additive relationship, 
and higher compression apparently 
does not result in sharp breaks in this 
relationship. The range of values of the 
specific volume of the Fe-S system in- 
cludes the seismically derived values of 
the core! The implication is that an 
iron-sulfur liquid composition is suffi- 
ciently dense and close packed to sat- 
isfy the currently assigned properties 
of the core. 

Hydrostatic compression of powders 
and crystals is generally instantaneous 
and reversible within a given phase re- 
gion. Hazen, Mao, Finger, and Bell 
have documented an exception to this 
rule in their studies of the pressure- 
volume relations of single-crystal wiis- 
tite (Fe^^O). Below about 150 kbar 
the crystals undergo a permanent vol- 
ume reduction. A crystal held at 200 
kbar for two weeks, for example, sus- 
tained a —0.8% volume change. Fur- 
thermore, volume reductions appear to 
be both time- and pressure-dependent, 
consistent with a diffusion-controlled 
process. A volume change of 0.8% cor- 
responds to a 3% reduction in the iron 
content of wiistite; one possible mecha- 
nism for this irreversible behavior is, 
thus, exsolution of iron metal from the 
wiistite structure. Reordering of wiis- 
tite vacancies and ferric iron or inter- 
action with the neon constituting the 
hydrostatic-pressure medium are also 
possible causes for this unusual phe- 
nomenon. In any event, irreversible 
compression may have introduced er- 
rors in previous static measurements, 
and a careful examination of samples 
before and after pressurization is thus 

The compressibility of wiistite 
(Fe !_<.()) was found by Hazen to vary 
systematically with the degree of non- 
stoichiometry. In a new procedure for 
measuring the relative, as well as the 
absolute, compressibility of single 
crystals, three wiistite samples of dif- 
ferent Fe content were placed in the 



same high-pressure mount. Volume 
changes of the three crystals were com- 
pared at several pressures to 55 kbar, 
and Fe .9oO was thus found to be 1.4 ± 
0.4% more compressible than Fe .947O, 
whereas Fe .93O has an intermediate 
value. These systematic changes in 
compressibility are not sufficient to ac- 
count for the large differences in ob- 
served bulk moduli measured by static 
vs. dynamic compression techniques. 

In the study of the behavior of acti- 
nide metals at high pressure, Benja- 
min, Zou, Mao, and Bell determined 
the behavior of thorium metal and 
uranium oxide to 650 kbar at 25 °C. 
The compressibility data on thorium 
may be useful for evaluating its distri- 
bution in the primordial earth's mantle 
and as a fractionation product in the 
upper mantle. The compressibility 
data on uraninite obtained in this 
study are also useful in formulating 
actinide element distribution models of 
the earth. In addition to the volume vs. 
pressure data, these investigators 
found a new phase change in uraninite 
from the cubic fluorite-type structure 
to what appears to be a smaller vol- 
ume, orthorhombic phase at approxi- 
mately 450 kbar. 

In their investigation of elements at 
high pressure, Mao, Zou, and Bell 
found a new phase of tellurium (Te) and 
three new phases of selenium (Se). Both 
Te and Se are semimetals, whose elec- 
trical resistances change from poor in- 
sulators to semiconductors at high 
pressure and finally to metallic conduc- 
tors at higher pressure. The crystal 
structures of the new phases have not 
yet been identified, but x-ray diffrac- 
tion d values were obtained to approxi- 
mately 300 kbar on Te and to approxi- 
mately 500 kbar on Se. 

Gallium arsenide (GaAs) has attracted 
attention because of its semiconductiv- 
ity at 1 bar. Previous high-pressure re- 
search suggested that gallium arsenide 
may be a fixed-point calibrant, because 
of a transition observed at approxi- 
mately 180 kbar and 25 °C. In coopera- 

tion with the U.S. National Bureau of 
Standards, several laboratories in the 
United States and abroad were asked 
to determine the pressure of the transi- 
tion for comparison. Bell and Mao ob- 
served the transition in a single crys- 
tal of GaAs, using neon as a hydro- 
static pressure-transmitting medium. 
At approximately 70-80 kbar, GaAs 
changed from opaque to translucent. 
This change indicated a shift of the ab- 
sorption edge and the band gap. Be- 
tween 180 and 190 kbar, a gradual 
change from translucent back to 
opaque was observed; again a shift of 
the absorption edge was suggested. 
The last change persisted as the pres- 
sure was released, an effect they believe 
is consistent with a phase change. 

Bell and Mao designed a multiply 
ported, diamond-window, high-pres- 
sure cell for determining seismic veloci- 
ties and elastic constants of fluid and 
small crystals at high pressures. They 
joined with Prof. E. Brody and Dr. H. 
Shimizu, who were at the Optics Insti- 
tute, University of Rochester, to study 
crystallized hydrogen and deuterium 
by the Brillouin-scattering spectro- 
scopic technique. They measured the 
velocities, density, and refractive index 
of the two hydrogen isotopes to a pres- 
sure of 200 kbar. The results will have 
bearing on studies of the equation of 
state of hydrogen and its application to 
models of planetary interiors. The ap- 
paratus used in conjunction with the 
pressure cell for making the Brillouin 
measurements is a laser interferom- 
eter, one of which has been constructed 
at the Geophysical Laboratory. Mea- 
surement of the elastic properties of 
other crystals in situ at pressures of 
over 100 kbar can now be realized in 
studies that relate laboratory data to 
seismic observations on the earth. 

In their study of hydrogen and deu- 
terium at high pressures and high tem- 
peratures, Bell, Mao, and Sharma 
noted a marked increase in the reaction 
rate between hydrogen (H 2 ) and deu- 
terium (D 2 ) as a function of pressure. 



The H 2 + D 2 - 2HD reaction is a 
classic in kinetic studies at 1 bar, and 
the effects of pressure are of special in- 
terest in studies of the Jovian planets. 
The equations of state of neon and 
argon, determined at high pressure last 
year by Hazen, Mao, Bell, and Finger, 
are particularly important for high- 
pressure studies because of the high 
compressibility of crystalline neon and 
argon. The data are especially relevant 
to the theory of materials at high pres- 
sure. This year, Zou, Finger, Hazen, 
Bell, and Mao extended the experi- 
ments to higher pressures and com- 
pared the experimental data with new 
quantum calculations. The new data 
were obtained at 25 °C and recalculated 
to K for comparison with theory. 
Highly improved values of the com- 
pressibilities of solid neon and argon 
were obtained, and it will be possible to 
calculate additional higher-tempera- 
ture isotherms. The substantial range 
of new data yields an equation of state 
of a solid material that provides a 
critical test of finite-strain theory. Such 
an equation provides new insights for 
the physicist on the repulsive forces in 
rare gas solids and provides a theoreti- 
cal base for geophysicists to interpret 
compressibility of minerals. 

Liquid and Glass Structures 

One of the principal characteristics 
of an igneous rock is that it forms in 
whole or in part from a molten condi- 
tion. In the past, the focus has been on 
the properties of the crystals formed in 
the liquid; yet the properties of the liq- 
uid are equally important in defining 
the equilibria. A major effort is being 
made, therefore, to characterize the 
liquid or its quenched "equivalent" 
glass. Most studies on the structure of 
silicate melts have been carried out on 
glass. In order to assess the extent to 
which glass structure reflects the melt 
structure, Seifert, Mysen, and Virgo 
carried out Raman spectroscopic stud- 
ies on several melt compositions to 

temperatures as much as 200 °C above 
the liquidi. They found that the to- 
pologies of the Raman spectra were not 
significantly affected by melting. The 
only difference appears in the so-called 
"defect bands" involving three-dimen- 
sional network units in the melts. The 
defect density may increase from per- 
haps 0.1% in glass to about 1% in the 
melt. They concluded, therefore, that 
the types of structural units in a glass 
and its melt remain the same. Their 
proportions may, however, change as a 
function of temperature. It is also 
likely that temperature-dependent ef- 
fects, such as Al 3+ distribution be- 
tween structural units, will be affected 
by transition into the melt region. 

Determination of the exact propor- 
tions of structural units in liquid is re- 
quired to formulate thermodynamically 
meaningful models of silicate melts. 
Such determinations may be accom- 
plished by measuring cross sections of 
Raman bands derived from specific 
structural units. The cross sections 
were obtained by preparing a sequence 
of compositions wherein the types and 
proportions of the coexisting struc- 
tural units change. Inasmuch as the 
area ratios of relevant Raman bands 
have finite uncertainties (about 5%), 
the Raman cross sections were deter- 
mined by a least-squares minimization 
procedure from a number of linear 
equations equal to the number of un- 
known units in the melts. Sufficient ad- 
ditional equations are obtained from 
mass balance. The method has been 
successfully used by Seifert, Mysen, 
and Virgo for compositions on the join 
CaO-Si0 2 . They found that when mea- 
sured against the theoretical number 
of bridging oxygens per silicon in the 
melts, the determined proportions of 
structural units are within 3% of their 
theoretical value. It was concluded, 
therefore, that the proportions of 
structural units in a silicate glass may 
be determined quantitatively from its 
Raman spectrum. 

Aluminum has an important network- 



forming role in the liquids of natural 
magmas. Mysen, Virgo, and Kushiro 
consider the distribution of aluminum 
between the several structural units to 
affect magma properties, such as vis- 
cosity and compressibility. They also 
believe that crystal-liquid trace ele- 
ment partition coefficients will be a 
function of Al distribution between the 
structural units. They have used Ra- 
man spectra of silicate melts on the 
joins CaAl 2 04-CaSi205 and NaA10 2 - 
Na 2 Si 2 5 to calculate the partitioning 
of Al 3+ as a function of Al content and 
Ca/Na in the melt. They found that 
Al 3+ always shows a preference for the 
most polymerized structural unit. The 
extent of this preference appears to in- 
crease with decreasing field strength of 
the charge-balancing cation. Further- 
more, for melts with less than about 10 
mole % A1 2 3 , all Al occurs in the 
three-dimensional network units. With 
greater Al contents, Al is also found in 
sheet and chain units in the melt. It re- 
mains to be seen whether the changes 
in value of viscosity, for example, when 
measured, correlate with the deduced 
changes in liquid structure. 

Because natural magmas may con- 
tain more than 50% three-dimensional 
network structural units, it is impor- 
tant to determine the details of their 
features as functions of the charge- 
balancing cation and Al/Si of the melt. 
As a result of Raman spectroscopic 
studies on the joins MgAl 2 4 -Si0 2 , 
CaAl 2 4 -Si0 2 , and NaA10 2 -Si0 2 , Sei- 
fert, Mysen, and Virgo concluded that 
three-dimensional network structures 
in an aluminosilicate melt consist of 
combinations of interconnected six- and 
four-membered rings. In melts with 
alkali metals to charge balance Al 3+ , 
there are two six-membered rings with 
random Si, Al substitution in each ring. 
There is a slight preference for Al 3+ in 
the ring with the slightly smaller T-O-T 
angle (and longer T-0 distance). Melts 
having alkaline earth metals for charge 
balance consist of coexisting six-mem- 
bered rings of essentially pure Si0 2 , 

four-membered rings of Al 2 Si 2 8 2_ 
compositions, and, at Al/Si greater 
than approximately that of anorthite 
composition, six-membered rings of 
A1 2 4 2_ composition. The proportions 
of these rings change as a function of 
Al/Si of the melt, but the ring composi- 
tions remain unaffected. In natural 
magma, M 2+ /Na + increases with in- 
creasing basicity; therefore, the pro- 
portion of Al-free, six-membered rings 
increases relative to the total abun- 
dance of three-dimensional network 

Virgo, Mysen, and Seifert have deter- 
mined the relationship between the 
oxidation state of iron and the structure 
of some of the geologically important 
silicate melts at 1 bar using 57 Fe Moss- 
bauer spectroscopy. Silicate glasses 
with 1-5 wt % Fe 2 3 were prepared in 
the systems Na 2 0-Si0 2 -Fe-0, K 2 0- 
Si0 2 -Fe-0, BaO-Si0 2 -Fe-0, SrO-Si0 2 - 
Fe-O, and CaMgSi0 4 -Si0 2 -Fe-0 by 
quenching melts in air at 1400°- 
1500°C. In all these systems, Fe 3+ / 
Fe 2+ is systematically related to values 
of nonbridging oxygens per tetra- 
hedrally coordinated cations (NBO/T), 
but discontinuities in the trends of the 
ferrous-ferric ratio occur at values of 
NBO/T that define distinct composi- 
tional ranges of the same coexisting 
anionic species in the melt. Within each 
of these compositional ranges, the 
Fe 3+ /Fe 2+ increases with increasing 
NBO/T except in Na 2 0, K 2 0, SiO, and 
BaO systems when less than 20 mole % 
metal oxide is present. The Fe 3+ /Fe 2+ 
also increases with increasing Fe 2 3 
content. At values of similar NBO/T, 
Fe 3+ /Fe 2+ is approximately the same in 
Na 2 0, K 2 0, SrO, and BaO melts but is 
significantly larger (a factor of two) in 
CaO and MgO melts. From the hyper- 
fine parameters of the Mossbauer 
spectra, Virgo, Mysen, and Seifert 
show that whereas ferrous iron occurs 
as a network modifier in all melts stud- 
ied, ferric iron is predominantly tetra- 
hedrally coordinated in Na 2 0, K 2 0, 
SrO, and BaO melts as M+IM 2+ Fe 2 4 



complexes. In CaO and MgO melts, 
however, Fe 3+ occurs principally as a 
network modifier. The Fe 3+ /Fe 2+ in 
these melts is directly proportional to 
the ratio of tetrahedrally to octahe- 
drally coordinated Fe 3+ . These results 
will be of importance in multicompo- 
nent melts, and Fe 3+ /Fe 2+ may, in 
principle, be determined by the relative 
stabilities of tetrahedrally coordinated, 
charge-balanced complexes. 

Both the physical and chemical prop- 
erties of a melt will be affected depend- 
ing on whether or not ferric iron is in 
tetrahedral coordination. As a network- 
former, ferric iron must be charge- 
balanced with Na + or K + , whereas 
Al 3+ may be charge-balanced with 
either alkali metals or alkaline earths. 
Inasmuch as most magmas are not per- 
alkaline, the conclusions of Mysen, 
Virgo, and Seifert may indicate that 
ferric iron can be only a network-former 
in magmas if alkali-ferrite complexing 
results in more stable structural units 
than the analogous aluminate complex. 
Their Mossbauer spectroscopic studies 
indicate that ferrite complexing is 
highly dependent on the proportions of 
A1 3 +,M+, andM 2+ . 

The oxidation state is also important 
to the roles in which other elements af- 
fect the characteristics of a melt. Virgo, 
Seifert, and Mysen have studied Eu 2+ / 
Eu 3+ equilibria in reduced silicate melts 
along the joins CaMgSi0 4 -Si0 2 -Eu-0 
and Na 2 0-Si0 2 -Eu-0 using 151 Eu 
Mossbauer spectroscopy. In alkaline- 
earth melts containing 5 wt % Eu 2 3 , 
Eu 2+ /Eu 3+ systematically decreases 
with increasing NBO/Si over the range 
1.3-3.2. The Eu 2+ /Eu 3 + also decreases 
at fixed NBO/T over the range 1-5 wt % 
Eu 2 3 in the melt. Significantly, the 
Eu 2+ /Eu 3+ is oxidized four times more 
in alkali melts than it is in alkaline- 
earth melts. In general, these effects are 
similar to those found for Fe 2+ /Fe 3+ 
equilibria in melts. The significantly 
different Eu 2+ /Eu 3+ equilibria in alkali 
versus alkaline-earth melts cannot be 
explained in terms of distinct coordi- 

nation polyhedra for the more oxidized 
state. Virgo, Seifert, and Mysen specu- 
late that the formation of a charge- 
balanced complex involving Na + and 
Eu 3+ in the cation matrix of the melt 
may lead to the increased stability of 
trivalent europium in alkali melts. 

Igneous Petrology 

The rock-forming oxide minerals are 
commonly lumped together by the 
petrographer as "opaques"; yet they 
contain valuable information regarding 
the processes and conditions of forma- 
tion of their host rocks. The chromium 
and iron oxides have been given special 
attention because of their growing im- 
portance both as strategic ore minerals 
and as unique indicators of P, T, f 02 , 
age, and rate of formation, as well as 
element saturation. 

Irvine presents a refined mechanism 
for chromitite formation in the Muskox 
Intrusion, based on melt densities cal- 
culated for a high-temperature, phase- 
equilibrium system and observations 
from model experiments on double- 
diffusive convection in aqueous solu- 
tions. The density data show that melts 
along the olivine- Ca-poor pyroxene- 
chromite primary phase boundary have 
almost constant density over a tem- 
perature range of more than 300 °C. 
This boundary is significant because 
Muskox chromitite layers are inter- 
stratified with cumulate layers of 
olivine and Ca-poor pyroxene. The im- 
plication of the density relationship is 
that melts at high- and low-temperature 
points along this boundary will mix 
readily, and Irvine shows from the 
phase-equilibrium data that such mix- 
ing should lead to crystallization of 
chromite by itself! That process con- 
stitutes a unique and significant alter- 
native view in our understanding of 
layered intrusions. Irvine also presents 
a physical model illustrating how this 
process might occur as the cumulate 
layers crystallize laterally from a den- 



sity-stratified column of convective liq- 
uid layers. The appreciation of these 
new concepts led Irvine, together with 
S. G. Todd, D. W. Keith, and other 
geologists of the Johns-Manville Cor- 
poration, to prepare a major paper on 
the J-M Platinum-Palladium Reef of the 
Stillwater Complex. In addition, the 
new interpretations led Irvine to revise 
the terminology for layered intrusions. 

Osborn and Boctor report on the 
significance of the large range in Cr 2 3 
content of magnetite in calc-alkaline 
extrusive rocks. They view the varia- 
tion in Cr 2 3 content as largely a func- 
tion of temperature and composition of 
the magma, in analogy with phase rela- 
tions for the system MgO-iron oxide- 
Cr 2 3 -Si0 2 . They conclude that the 
Paricutin succession of andesite flows 
represents a series of magmas formed 
at decreasing temperatures. The Cas- 
cades calc-alkaline magmas, as repre- 
sented by specimens of high-Al basalt 
and andesite, show a similar but less 
pronounced temperature decrease from 
basalts to high-Si0 2 andesites; and the 
Santorini magmas, ranging from 50.8 
to 54.5% Si0 2 , were all at the low tem- 
perature indicated for the high-Si0 2 
andesites of the Cascades and the late 
(1951) Paricutin andesite. 

The opaque minerals in nodules from 
kimberlite dikes have also been studied 
by Boyd, Nixon, and Boctor, however, 
their immediate interest was in some 
unusual textures in a suite of pyrox- 
enite xenoliths. Pyroxenites from the 
Mzongwana kimberlite dike, Transkei, 
have quench textures including spher- 
ulitic growths and mineral segrega- 
tions believed to have formed by rapid 
crystallization of supersaturated melts. 
Coexisting garnet and amphibole in 
the pyroxenites are evidence that they 
originated at a depth near 100 km, 
where they were entrained by kimber- 
lite erupted from a depth of the order of 
150 km. Boyd, Nixon, and Boctor be- 
lieve that the quench textures formed 
when globules of liquid pyroxenite 
magma were included in the erupting 

kimberlite and were quenched by ex- 
pansion of the fluidized kimberlite. 

Of considerable importance in the 
transport of such magma is the viscos- 
ity of liquid. Viscosities of liquids of 
basalt, andesite, and obsidian decrease 
with increasing pressure under isother- 
mal conditions. In all these melts the 
ratio of nonbridging oxygens (NBO) to 
tetrahedrally coordinated cations (T) is 
less than 1 (i.e., NBO/T < 1). Con- 
versely, it was believed that in silica- 
deficient magmas with NBO/T > 1 
(such as basanite, picrite, and koma- 
tiite) there would be an increase in 
viscosity with increasing pressure. To 
examine the relationship of viscosity to 
liquid structure, Scarfe chose three 
compositions for study up to 15 kbar. 
An alkali olivine basalt composition 
(NBO/T = 0.63) decreased about 60% in 
15 kbar; sodium melilite (NBO/T = 0.67) 
decreased about 30%; and olivine ne- 
phelinite (NBO/T = 0.75) decreased 
about 20%. However, the uncertainty 
of these experiments is ±20%. In that 
pressure range the densities of the melt 
increase so that the kinematic viscosity 
decreases at an even greater rate with 

Because of the special role of alu- 
minum in forming complexes that in- 
fluence the viscosity of a magma, 
Kushiro measured the viscosities along 
the CaAl 2 4 -Si0 2 join by the falling- 
sphere method. He found that the rate 
of decrease of viscosity with pressure is 
reduced with increasing CaAl 2 4 com- 
ponent in the melt. This behavior dif- 
fers from that of melts on the join 
NaA10 2 -Si0 2 where the pressure de- 
pendence of viscosity increases with in- 
creasing A1/(A1 + Si) of the melt. 
Kushiro concluded that the viscosity 
decrease of anhydrous melts of natural 
basalt and andesite compositions up to 
at least 15 kbar may be due to both a 
weakening of the T-O-T bands with in- 
creasing aluminum content and a de- 
crease in ring size of the polymerized 
species in the melt deduced by Seifert, 
My sen, and Virgo (this Report). It is 



evident that the parameters influencing 
the speciation of melt have an impor- 
tant effect on its viscosity. 

In order to resolve some of the 
discrepancies in the phase diagram of 
LiAlSi0 4 -Si0 2 -H 2 0, London used wa- 
ter-clear and colorless natural minerals 
in the presence of excess quartz and 
controlled amounts of H 2 to avoid the 
leaching problem. He confirmed that 
petalite (LiAlSi 4 O 10 ) may break down 
to eucryptite (a-LiAlSi0 4 ) + quartz with 
decreasing temperature, as these three 
minerals occur in some lithium-bearing 
pegmatites. The rarity of this assem- 
blage, restricted to low pressures and 
temperatures, supports the view that 
spodumene (a-LiAlSi20 6 ) and petalite 
are indeed the principal and primary 
lithium aluminosilicates in pegmatites. 

Some 80,000 rock analyses probably 
exist in the literature, but little more 
than a third of these are accessible elec- 
tronically for penological investigation. 
It is perhaps optimistic to think that 
all chemical analyses of rocks will some 
day be available in this form. Never- 
theless, there are sufficient data to test 
in a preliminary way many petrologic 
theories. Chayes noted many years ago 
that expressing the chemical analysis 
of a rock in the form of percentages 
automatically generates a strong bias 
toward negative correlation among the 
more variable components in any set of 
proportions, essentially invalidating 
coventional statistical procedures for 
testing the significance of association 
between pairs of such variables. The 
inferential problem created by the 
"closure constraint" was also noted 
about the same time by Russian geolo- 
gists and American biometricians. It 
has been the subject of much stimulat- 
ing research wherein it is supposed that 
random samples drawn from an "open" 
parent of independent variables are 
"closed," either by nature or by the 
observer, into a set of proportions of 
percentages. A second transformation, 
referred to by Chayes as a "remaining- 

space transformation," may then be 
invoked as a means of reducing or elim- 
inating the biasing of covariances aris- 
ing from the initial closure. It is in fact 
a further closure, or projection, entirely 
excluding one of the variables of inter- 
est. The parent open distribution used 
in theoretical work to date, however, is 
one in which variances are proportional 
to means. Chayes has now shown that 
if that proportionality exists, the ranks 
of closed variances and means will be 
identical. If they are not, and in some 
of the penological data used to illus- 
trate discussions of the various remain- 
ing-space tests they are very far from 
identical, the observed sets of propor- 
tions could not have resulted from clo- 
sure of a parent distribution of the pro- 
posed type. It is, therefore, unneces- 
sary to inquire whether its covariances 
could have been generated in this fash- 
ion. A quick glance at the data may ac- 
complish as much in this respect as a 
rather elaborate set of calculations. 
More importantly, however, the rank 
relations of means and variances in 
most Harker arrays render the results 
of such tests uninterpretable. 

* Design and development of a world 
data base for igneous petrology con- 
tinues as a major component of the 
research of Chayes, who is chairman of 
an International Geological Correlation 
Project supervising this work. Under 
his leadership, the project's U.S. group 
embarked on a systematic scan of the 
major serial and periodical source 
references published in the United 
States since 1917. He estimates that 
this scan is now about 40% complete; 
all data accumulated to date have been 
transferred from coding sheets to 
machine-readable files, and galleys of 
the files have been returned to con- 
tributors for proofreading. The interna- 
tional organization stimulates and 
monitors the work of a considerable 
corps of volunteers, who do the actual 
literature scan; by far the most prob- 
lematic step in the construction of the 



base, then, is the transfer of data from 
the final hard copy to machine-readable 
form. The long-desired decentralization 
of data-transfer work has begun to 
materialize, and Chayes anticipates 
that a number of U.S. contributors will 
soon begin submitting data in machine- 
readable form rather than on coding 
sheets. This accomplishment brings 
the project to the state at which inten- 
sive work on the structure of the base 
and its associated information system 
must be resumed, and such work is now 
under way. 

*The data transfer is complex be- 
cause of the varying levels of com- 
pleteness of rock description. A pro- 
gram has now been devised by Chayes 
so that advanced students working 
under professional supervision can pro- 
duce data transfer efficiently. Proce- 
dures for recording trace amounts and 
radiochemical ages have been rational- 
ized and standardized, as have those 
for framing and tagging unformatted 
information of various types. Existing 
machine-readable data files have been 
edited to conform to these new stan- 
dards, and they will be used in all fu- 
ture transfer work. Use of these stan- 
dards will greatly facilitate translation 
in the event that major changes in in- 
put format seem desirable. 

♦Effective exploitation of a base of 
the size and type contemplated will re- 
quire sophisticated search procedures. 
With the objective of minimizing 
search costs to the ultimate user, 
Possolo has implemented, for a small 
trial set of parameters, a restructuring 
corresponding to a collection of "in- 
verted linked-lists." The reduction of 
execution time by inversion must be 
paid for by increased storage re- 
quirements, and if the number of keys 
is large the increase in storage require- 
ment will also be large. The most prac- 
tical way to discover the balance be- 
tween reduced execution charges and 
increased storage charges appears to 
be by trial operation of a pilot base. 

Element Partitioning 

Trace element and isotopic data from 
ultramafic nodules in kimberlite and 
alkali basalt indicate that the source 
regions of these rocks may have under- 
gone metasomatic alteration. Mysen 
has performed rare earth element (REE) 
partitioning experiments to determine 
crystal-fluid partition coefficients rele- 
vant to such processes. The experimen- 
tal results show that even for pure C0 2 
fluids, the REE will have a preference 
for the fluid over any other phase in the 
upper mantle. The pressure effect is di- 
minished, however, as the C0 2 content 
of the fluid increases. Mysen showed 
that fluid-mineral partition coefficients 
increase rapidly with decreasing pres- 
sure under isobaric conditions and that 
the temperature effect increases with 
increasing C0 2 /H 2 of the fluid. In ap- 
plying these data to rock-forming pro- 
cesses, he concluded that most likely 
the source rock of alkali basalt had un- 
dergone C0 2 -rich metasomatism prior 
to partial melting at a depth equivalent 
to about 30 kbar. The source of the meta- 
somatizing fluid could be related indi- 
rectly to the source rock of the alkali ba- 
salt. Alteration of the ultramafic rock 
from which the nodules in kimberlite 
were derived probably occurred at pres- 
sures in excess of 50 kbar. The source of 
the fluid apparently was eclogitic; it 
could not have been a garnet lherzolite, 
as the fluid from such a rock would not 
show sufficient light REE enrichment. 

As suggested above, peridotitic source 
rocks for alkaline magmas may have 
been subjected to metasomatic pro- 
cesses by fluids, as a precursor to 
magma generation. Amphiboles from 
these rocks have F contents greater 
than those of the more common composi- 
tions found in the majority of mantle- 
derived rocks. It is likely that F was an 
important constituent in such metaso- 
matic fluids and altered the element 
partitioning between silicates and va- 
por. As a first step in determining the 



behavior of F-bearing vapor, Danck- 
werth has obtained preliminary results 
for the system albite-HF-H 2 at 15 
kbar. In addition to the depression of 
the liquidus, fluorine apparently is 
preferentially partitioned into the liq- 
uid and increases the incongruent na- 
ture of the partitioning of silicates into 
the vapor. For example, in experimental 
charges with HF/(HF + H 2 0) > 10%, 
quartz as well as albite appeared at 
600 °C and 15 kbar. 

The partitioning of Ni between fayal- 
ite and iron monosulfide was previously 
studied at a subsplidus temperature of 
900 °C, and fugacities of oxygen and 
sulfur were specified by the quartz-fay- 
alite-magnetite-pyrrhotite buffer. In 
the present investigation by Boctor, Ni 
partitioning between olivine solid solu- 
tions (Fo 82 -Fo 87 ) and iron monosulfide 
was studied in the temperature range 
1300° -1400 °C. Gas mixtures of H 2 , 
C0 2 , and S0 2 were used to control the 
/o 2 and /s 2 m the experiments. Nickel 
partitioning was found to favor the sul- 
fide phase relative to olivine regardless 
of temperature, olivine composition, or 
Ni concentration. The data can be used 
in interpreting the origin of iron-nickel 
sulfide ores of the Sudbury type and of 
those associated with komatiite flows. 
The data also support the suggestion 
that the separation of an immiscible 
sulfide liquid during the ascent of 
tholeiitic magmas would lower the Ni 
content to the levels observed in basal- 
tic rocks. 

The univariant assemblage pyrite- 
pyrrhotite-magnetite in the system Fe- 
O-S occurs in many ore deposits of 
magmatic and metamorphic origins. 
The partitioning of Ni and Co between 
these phases was studied by Boctor in 
experiments with evacuated silica 
tubes in the temperature range 300°- 
600 °C. The partitioning of Ni favors 
pyrrhotite relative to pyrite and mag- 
netite, whereas Co partitioning favors 
pyrite relative to the other two phases. 
The partitioning of Ni and Co between 
pyrite, pyrrhotite, and magnetite is 

temperature dependent; therefore, 
these phases can be used as potential 
geothermometers in those ores in 
which they occur as an equilibrium as- 

Fluid Speciation and Dynamics 

Perhaps one of the most critical areas 
for research today is in ascertaining the 
species in an element-transporting fluid. 
Furthermore, it is essential to know, on 
both a local and a regional scale, the di- 
rection of flow of that fluid. Such 
knowledge underlies the entire quest 
for ore deposits and the understanding 
of the processes of metamorphism. As 
is the custom at the Geophysical Labo- 
ratory, the principles are first worked 
out with simple systems in which the 
variables can be defined and controlled. 
The principles are then applied to field 
observations to determine whether 
other variables should be taken into 
account. Below are described three dif- 
ferent approaches to the fluid specia- 
tion and dynamics problem. 

Mineral-solution equilibria in the sys- 
tem MnO-Si0 2 -HCl-H 2 were studied 
by rapid-quench hydrothermal tech- 
niques with the Ag-AgCl buffered and 
unbuffered methods at 2 kbar in the 
temperature range 400° -700 °C. In this 
temperature and pressure range, 
MnCl 2 ° seems to be the predominant 
Mn-bearing species in the fluid with 
rhodonite and quartz. Data at low tem- 
perature, though suggestive of ap- 
preciable association, are inconclusive. 
The equilibrium constant for the reac- 
tion involving rhodonite, quartz, and 
HC1 was determined by Boctor and 
Frantz in the temperature and pressure 
ranges of the experiments and was 
used to calculate the free-energy dif- 
ference between MnCl 2 ° and HC1°. The 
data constitute part of the information 
needed to calculate the fluxes of metals 
in hydrothermal systems associated 
with mid-oceanic ridges. The data can 
also be used to calculate the solubility 
constants of Mn minerals for which 



thermochemical data are available, and 
to predict solution compositions in 
equilibrium with mineral phases in Mn- 
rich skarns. 

Frantz and Marshall have measured 
electrical conductances of calcium chlo- 
ride and magnesium chloride in dilute 
aqueous solutions from 25° to 600 °C at 
pressures up to 4 kbar. The high-tem- 
perature conductance and association 
behavior of these two salts, present in 
natural waters, are of fundamental inter- 
est in understanding geothermal sys- 
tems. Solution compositions of the 
salts used were between 0.001 and 
0.005 molal. Frantz and Marshall used 
the results of these experiments, re- 
ported in Year Book 78, to compute the 
two stepwise ionization constants of 
both salts. The logarithms of these con- 
stants were found to vary linearly with 
the logarithms of the water density and 
reciprocal temperature. The first ioni- 
zation constants of both salts were 
found to be within half an order of mag- 
nitude of those for sodium chloride. It 
is now possible to calculate the distri- 
bution of magnesium chloride and cal- 
cium chloride species in equilibrium 
with silicate minerals in hydrothermal 

Rumble and Hoering have deter- 
mined that infiltration rather than in- 
tergranular diffusion through static 
pore fluid is the most effective mecha- 
nism of oxygen isotope equilibration 
during metamorphism. They measured 
the 6 18 of quartz separated from 
small samples spaced at 1-cm inter- 
vals across contacts between rocks 
whose 18 content differed signifi- 
cantly prior to metamorphism. In out- 
crops with no evidence of fluid flow 
during metamorphism, differences in 
composition of 1 %o have been found in 
quartz separated from adjacent sam- 
ples 1 cm apart. Quartz from outcrops 
with evidence of fluid flow, such as in- 
filtration metasomatic mineral zones, 
however, is homogeneous over dis- 
tances of several meters. The data con- 
firm that infiltration is a mechanism 

capable of transferring oxygen isotopes 
over distances greater than 1 m, and it 
is probably the most efficacious mecha- 
nism for oxygen isotope equilibration 
during metamorphism. 

Comparative Crystal Chemistry 

Understanding of the physical and 
chemical behavior of a crystal depends 
on accurate knowledge of its structure. 
The interpretation of all properties 
eventually must be related to the in- 
teratomic forces in the crystal. The 
crystallographers at the Geophysical 
Laboratory have provided the funda- 
mental structural information on which 
to base other types of investigations 
seeking characterization of those in- 
teratomic forces. One of the most direct 
contributions has been in ascertaining 
the compositional limits of a crystal im- 
posed by its general structural arrange- 
ment under a given set of conditions. 
For example, kornerupine, (Mg, Fe, 
Al) 9 _ 10 (Si, B, Al) 5 (O, OH) 22 , is a com- 
plex silicate that, unlike many miner- 
als, occurs with limited substitution of 
iron for magnesium. Finger and Hazen 
have refined the crystal structure and 
determined the site ordering of a sam- 
ple with Fe/Mg = 0.5, a composition 
near the most iron-rich natural sample. 
Iron and magnesium are disordered 
over three of the seven octahedral and 
larger sites. There are no obvious struc- 
tural limitations on the maximum iron 
content, at least under the conditions 
of room temperature and pressure. The 
mineral coexists with other iron- 
bearing phases such as biotite, cor- 
dierite, garnet, and possibly alkali 
feldspar, so it would appear that the 
iron content of kornerupine is governed 
by the partitioning coefficients rather 
than the structure. In short, the rela- 
tive free energies of all sites in which Fe 
may exist in all coexisting phases de- 
termine the population of iron ions in 
each site. The calculation of the abso- 
lute free energy of each site, from which 
the relative free energies may be evalu- 



ated, is indeed one of the long-range 
goals of the crystallographer, the ther- 
mochemist, and the solid-state physi- 

Special apparatus is required to as- 
certain the constraints on composition 
in a crystal as a result of its structure 
under the conditions of formation. A 
high-temperature, diamond-anvil cell 
(PT cell) has been developed for struc- 
ture determinations under geologically 
relevant conditions. The studies on 
diopside and orthoenstatite constitute 
the first single-crystal data at com- 
bined elevated temperature and pres- 
sure for any silicate. These are the sec- 
ond and third single crystals to be x- 
rayed in such an environment. 

The average density of rocks in- 
creases with depth because of phase 
transitions and compositional varia- 
tion, but the average density of a 
given mineral assemblage may be 
greater at the earth's surface than in 
the upper mantle. Structures of diop- 
side (CaMgSi 2 6 ) have been deter- 
mined by Hazen and Finger under 
three PT conditions: 17 kbar, 110°C; 21 
kbar, 180°C; 13 kbar, 375 °C. Lines of 
constant diopside unit-cell volume 
have a slope of 40 bar/°C, which is more 
than twice the average geothermal gra- 
dient. Thus diopside, as well as many 
other dense silicates, may be less dense 
at depth within the earth than at the 
surface. The structure of diopside var- 
ies along an isochore (PT line of con- 
stant volume) because the large M2 
calcium site expands relative to the 
rest of the structure at high PT. This 
behavior is probably due to rapid ex- 
pansion of the long M2-03 bond. 
Three-dimensional structure refine- 
ments at PT yield reasonable tempera- 
ture factors, but bond distance errors 
average ±0.01 A; this precludes unam- 
biguous comparison of the subtle dif- 
ferences of isochoric structures. 

Olivine and orthopyroxene are the 
principal minerals thought to compose 
the upper mantle of the earth. Their 
compositions are near those of the 

magnesian end members of their re- 
spective solid-solution series. Ralph, 
Hazen, and Finger have investigated 
the structure of orthoenstatite in the 
PT cell at pressures up to 25 kbar and 
temperatures to 285 °C, and they have 
determined cell parameters of moder- 
ate precision when compared to am- 
bient measurements. When plotted 
against normalized unit-cell volume, 
axial ratios do not differ significantly 
from those measured under high pres- 
sure alone. The ratios show linear 
trends, suggesting that the modes of 
structural response to temperature and 
pressure are similar within the range of 
conditions studied. Structure refine- 
ments of orthoenstatite from intensity 
data collected in the PT cell (and in an 
unmodified pressure cell) have not yet 
yielded acceptable results, perhaps 
because the crystal is in a metastable 
region. Studies of orthoferrosilite, the 
isostructural ferrous orthopyroxene 
end member with inherently more in- 
tense x-ray scattering than orthoensta- 
tite, are now in progress and may yield 
a more acceptable refinement. 

The crystal structure of the new min- 
eral bannermanite, NaV 4+ V 5+ 5 15 , 
has been refined by Hughes, Finger, 
and Hazen. This material, formed in a 
high-temperature fumarole at Izalco 
volcano, El Salvador, is the only known 
natural "oxide bronze." This phase is 
one of the compounds with the general 
formula M y TO n , where M is an elec- 
tropositive metal ion and T is a transi- 
tion metal ion. The M ions donate elec- 
trons to the TO n subarray with charge 
balance maintained by reduction of one 
transition ion per donated electron. In 
bannermanite, the M cations are 7- 
coordinated sodium and the T ions are 
vanadium. In this structure Na positions 
related by symmetry have a separation 
of only 1.81 A, an impossibly short 
distance. The structure cannot occur, 
therefore, with more than 50% occu- 
pancy of that site; hence the composi- 
tional range is limited. 

Pinckney, Finger, Hazen, and Burn- 



ham determined the high-temperature 
crystal structures of the 7-repeat py- 
roxenoid pyroxmangite (MnSi0 3 ), as 
part of an ongoing study of relation- 
ships between structure and stability 
of metasilicates. Silicon-oxygen bond 
distances did not change significantly 
with temperature, whereas all Mn-0 
bonds displayed appreciable expan- 
sion. Furthermore, long Mn-0 bonds ex- 
panded faster than short bonds. Unit- 
cell parameters measured at high pres- 
sure mirrored changes that occur at 
high temperature. Pyroxmangite is 
known to transform to rhodonite at 
high temperature, and although this 
transition did not proceed to comple- 
tion, a slight splitting of diffraction 
maxima above 400 °C may be due to in- 
cipient coherent growth of rhodonite 
along (001) planes. The observed dif- 
ferential expansion of Mn-0 versus Si- 
O, and the consequent misfit of adja- 
cent cation polyhedra, appears to be a 
key to the stability of pyroxmangite 
and its polymorphs. 

Bio geochemistry 

An exciting area for research today is 
in the application of biochemistry to 
earth science problems. Paleontology 
has been expanded by the discovery of 
residual, indigenous, organic material 
in fossils. Stratigraphy has been given 
new tools for correlation, even where 
index fossils are absent or morphologi- 
cally indeterminate. The petroleum in- 
dustry has received benefits, and eco- 
nomic geology has acquired new insights 
from the interactions, passive or active, 
of ore solutions and organic material. 
Biochemistry itself gains from the new 
techniques demanded by earth scien- 
tists for dealing with minute amounts 
of complex organic assemblages. 

The new field resulting from the in- 
teraction of biochemistry with geology— 
biogeochemistry— is at that pioneering 
stage where every experiment is a new 
challenge and yields a new array of inter- 
esting questions. Reports of D-enantio- 

mers of amino acids in diseased tissue, 
for example, suggest that the techniques 
developed for biogeochemical investi- 
gations will eventually help resolve 
some health problems. 

Because workers at the Geophysical 
Laboratory enjoy freedom to focus on 
any substantive problem without re- 
gard to discipline boundaries, it should 
not be surprising that a current investi- 
gation involves thermophilic algae and 
photosynthetic bacteria. Organisms 
similar to those important in the forma- 
tion of Precambrian stromatolites 
(domal, laminated, calcareous sedimen- 
tary structures) are found today in the 
thermal springs of the western United 
States. These springs are inhabited by 
mat-forming, blue-green algae and fila- 
mentous photosynthetic bacteria. Estep 
has measured the stable carbon and hy- 
drogen isotopic compositions of several 
mats collected from hot springs in Idaho. 
Blue-green algae have a distinctly dif- 
ferent isotopic composition from the 
photosynthetic bacterium, Chloro- 
flexus, which at higher temperatures 
(55 °C) forms pure mats. When Chloro- 
flexus grows in association with uni- 
cellular blue-green algae, however, its 
isotopic composition shifts to one ap- 
proaching that of the blue-green algae. 
This shift indicates a change in the bac- 
terium's metabolism from photosynthe- 
sis to photohetero trophy; in essence, 
the bacterium grows by using photosyn- 
thetically produced organic matter 
from the blue-green algae. Even though 
the organic carbon and bonded hydro- 
gen can be transferred from one species 
to another, the isotopic composition of 
these algae and bacteria cannot be 
equated with that of the Precambrian 
stromatolites. In addition, no known 
nonbiological process produces the re- 
quired shift in isotopic composition, so 
other biological causes must be sought. 

One of the major accomplishments of 
the past few years— the application of 
geochemistry to stratigraphic prob- 
lems—has provided new tools to sup- 
plement classical methods. Corrado 



and Hare have performed ion-ex- 
change, amino acid analyses on approx- 
imately 180 individuals of the pelecypod 
Mulinia lateralis (Say) from Pleistocene 
sediments in the Coastal Plain of South 
Carolina near Charleston and Myrtle 
Beach. In each region, groupings of al- 
loisoleucine/isoleucine values were in- 
dicative of relative geologic age. Strati- 
graphic units in Charleston and Myrtle 
Beach have been tentatively correlated 
on the basis of alloisoleucine/isoleucine 
value groupings, ranges in elevation of 
associated Coastal Plain terraces, and 
geomorphic evidence. The number of de- 
positional events that can be inferred 
from the value groupings and the strik- 
ing similarity of the groupings them- 
selves suggest a correlation of strati- 
graphic units between Charleston and 
Myrtle Beach. The ranges in elevation 
of the terrace features associated with 
the four value groupings in Charleston, 
from youngest to oldest, overlap al- 
most perfectly with ranges in elevation 
of the terraces associated with the 
youngest to oldest value groupings in 
Myrtle Beach. The sedimentary se- 
quence, in which the next-to-youngest 
fossiliferous unit in Charleston and 
Myrtle Beach is buried in an almost 
identical fashion by the youngest unit, 
again suggests a direct correlation of 
stratigraphic units between these local- 
ities. With paleo-amino acid analyses, 
it will now be possible to detect regions 
in the southern Atlantic Coastal Plain 
that have undergone faulting and, 
more generally, to correlate beds in 
which the fossils are not diagnostic or 
where the fossil assemblages record 
similar repetitive transgressive-regres- 
sive events. 

Organic matter preserved in fossils 
and rocks consists of extremely com- 
plex mixtures, and extensive separa- 
tions and purification must be per- 
formed before characterization of the 
units is possible. Hoering and Freeman 
have explored the potential of a new mo- 
lecular sieve, silicalite, for selectively 
absorbing and separating hydrocarbon 

molecules from petroleum and rock ex- 
tracts. Hoering exploited this tech- 
nique for isolating slightly branched 
hydrocarbon molecules from petrole- 
ums and rock extracts. Such molecules 
have been known before, but because of 
the lack of a suitable separation tech- 
nique, have never been characterized 
carefully. Hoering discovered that 
suites of these molecules in young 
sediments are relatively simple, con- 
sisting primarily of the 2-methyl and 
3-methyl substituted isomers of straight- 
chained hydrocarbons, the ratio of the 
two isomers alternating with increas- 
ing carbon number. In older, more 
mature assemblages of organic matter 
from ancient sediments, the distribu- 
tion is more complex, with 4-methyl, 
5-methyl, and 6-methyl isomers pres- 
ent. The ratio of the 2-methyl to the 
3-methyl isomer, however, increases 
regularly without alternation as a func- 
tion of the carbon number. Hoering 
postulates that the branched lipids of 
living organisms are the precursors to 
the 2-methyl and 3-methyl isomers. 
With increasing age or temperature of 
a sediment, extensive rearrangement 
and cracking of molecules give rise to 
the complex mixtures found in older 
rocks. The new analytical method will 
be valuable as a tool for characterizing 
hydrocarbon assemblages and will facil- 
itate laboratory simulation experi- 
ments designed to elucidate the mecha- 
nisms of molecular rearrangement of 
organic matter in sediments. 

With the exception of glycine, all the 
common protein amino acids can exist 
as nonsuperimposable mirror images 
called enantiomers. Investigation of 
the interconversion (racemization) of 
L-amino acids, the common protein con- 
stituents of living organisms, to D-amino 
acids (eventually forming equilibrium 
or racemic mixtures) has become an im- 
portant area of study for low-tempera- 
ture geochemists and biochemists. 
Racemization reactions have already 
been applied to a wide variety of studies, 
including stratigraphic correlation, nu- 



tritional biochemistry, and chemotax- 
onomy. The resolution of mixtures of 
most protein amino acid enantiomers 
has now been accomplished with liquid- 
chromatographic methods. 

* Using high-performance liquid chro- 
matography, Engel and Hare have 
investigated the effects of several vari- 
ables, in order to improve the resolution 
of mixtures of the d- and L-enantiomers 
of the basic amino acids. The effects of 
ionic strength, pH, and temperature 
proved to be important. Increasing the 
pH of the mobile phase to 7.1 and ionic 
strength to 0.6 M, and lowering the col- 
umn temperature to 0°C resulted in 
baseline resolution of the d- and 
L-enantiomers of lysine, histidine, and 
arginine. The flexibility of this tech- 
nique ensures adequate resolution of 
basic amino acid enantiomers. 

Success in resolving the basic amino 
acids prompted Engel and Hare to work 
out the procedures for resolving mix- 
tures of acidic, neutral, and basic amino 
acids. Gas chromatographic resolution 
of the d- and L-enantiomers of four 
amino acids ubiquitous in nature— tryp- 
tophan, lysine, arginine, and histidine— 
has for the most part been unsuccessful, 
primarily because of the low vapor pres- 
sures of these amino acid derivatives. A 
new procedure that employs a combina- 
tion of high-performance liquid chro- 
matographic and gas chromatographic 
techniques has been developed that can 
successfully resolve most of the com- 
mon protein amino acids. Application of 
this method for determining the extent 
of racemization of arginine, lysine, histi- 
dine, and tryptophan will provide new 
information about the kinetics of amino 
acid racemization in geological and bio- 
logical systems. 

Hydrolysis is the initial step in deter- 
mining the extent of racemization of 
amino acids in peptides and proteins. 
Hydrochloric acid, which is commonly 
used as the hydrolysis agent, invariably 
causes tryptophan to decompose. Engel 
has shown that substitution of mercap- 
toethanesulfonic acid for HC1 hydro- 

lyzes peptides without destroying or 
racemizing tryptophan. This hydrolysis 
method has been successfully employed 
in conjunction with the new chromato- 
graphic procedure of Engel and Hare to 
isolate and resolve the d- and L-enanti- 
omers of tryptophan in peptides. 

The extent of racemization of many 
acidic and neutral amino acids in fossil 
shells has been studied, whereas the 
basic amino acids (lysine, histidine, and 
arginine) and tryptophan have been, for 
the most part, overlooked. Engel and 
Hare have developed analytical pro- 
cedures that can be used to isolate and 
resolve the d- and L-enantiomers of 
arginine, lysine, histidine, and tryp- 
tophan in fossil and modern peptides 
and proteins. The new procedures have 
been successfully employed to deter- 
mine the occurrence and extent of 
racemization of these amino acids in a 
Pleistocene mollusk shell, Saxidomus 
giganteous, from Willapa Bay, Pacific 
County, Washington. Additional ex- 
periments are being planned to evaluate 
the kinetics of racemization of basic 
amino acids in fossil and modern shells. 

In another application of the new 
chromatographic procedures, Engel 
and Hare resolved the basic amino acid 
enantiomers in a-melanocyte stimulat- 
ing hormone. Mild heat treatment of a- 
melanocyte stimulating hormone (a- 
MSH) in the presence of alkali increases 
its biological activity. This increase has 
been attributed to partial racemization 
of the amino acid constituents of the hor- 
mone. It has been speculated that if the 
principal sites of racemization of alkali- 
heat treated a-MSH could be deter- 
mined, then substitution of one or more 
D-amino acids in these highly racemized 
positions might further enhance the bio- 
logical activity of this hormone. Substi- 
tution of D-phenylalanine for L-phenylal- 
anine in a synthetic analog of a-MSH 
resulted in the synthesis of the most bio- 
logically active a-MSH analog known to 
date. The new chromatographic proce- 
dures developed by Engel and Hare have 
resulted in the determination of the ex- 



tent of racemization of all the amino acid 
constituents of a-MSH heat-treated in 
the presence of alkalies. Arginine and 
histidine were found to be highly race- 
mized. Engel and Hare surmise that the 
substitution of D-His, D-Arg, or both, in 
a-MSH might result in other potent 
MSH analogs with prolonged biological 

*In still another application of the 
new chromatographic procedures, En- 
gel and Hare determined the distribu- 
tion and extent of racemization of 
amino acids in a carbonaceous meteor- 
ite. Uncertainty with respect to the 
origin of these compounds is primarily 
a result of the difficulty in ascertaining 
whether the amino acids are extrater- 
restrial. Racemic isovaline has been 
detected in carbonaceous meteorites, 
yet isovaline has never been reported 
as a constituent of terrestrial biological 
systems. Any information that can be 
obtained, therefore, about isovaline 
should be useful for evaluating the 
origin and distribution of amino acids 
in extraterrestrial samples. With the 
new system, isovaline can be isolated 
rapidly, and its enantiomers can be 

*With the new techniques, it ap- 
peared useful to Engel and Hare to 
reexamine what effects time and 
temperature have on the extent of 
racemization of amino acids during the 
acid hydrolysis step, especially on those 
in the basic group that have not been 
previously studied. They determined 
the extent of racemization of arginine, 
histidine, tryptophan, and lysine during 
hydrolysis with 6 N HC1 at 102°, 
110.5°, and 150°C for 24, 22, and 0.25 
hr, respectively. A high concentration 
of tryptophan was used to compensate 
for its decomposition during 6 N HC1 
hydrolysis. Whereas these amino acids 
were, as expected, only slightly race- 
mized during hydrolysis, the condi- 
tions employed caused slight but uni- 
form differences in the rates of amino 
acid racemization. For example, all the 
amino acids were more racemized when 

heated at 110.5°C for 22 hr than at 
102 °C for 24 hr or 150°C for 0.25 hr. 
Experiments are also being conducted 
to determine the extent of racemization 
of these amino acids during acid hy- 
drolysis with mercaptoethanesulfonic 

New Techniques and Methods 

In recent experiments with the 
diamond-window, high-pressure cell, 
Bell and Mao have employed crystal- 
lized argon, neon, and helium as pres- 
sure-transmitting media. The crystal- 
line forms of these gases are weak solids 
in certain pressure ranges and intro- 
duce little or no error due to stress dif- 
ferences on a sample. The pressure is al- 
most hydrostatic with these media, and 
thus anisotropic, nonhydrostatic, and 
inhomogeneous stress effects that re- 
sult in errors, which are difficult to as- 
sess, can be avoided. Argon was used to 
110 kbar, and neon to 210 kbar, before 
finite strength was sufficient to affect 
the stress distribution transmitted to 
the sample. Helium was tested to 600 
kbar and presumably could be used at 
higher pressures. A method to generate 
hydrostatic conditions in ultrahigh- 
pressure experiments (pressure above 
100 kbar) has not been available before. 
Because the errors of nonhydrostatic 
stresses in past experiments could be 
large, especially in the relatively strong 
and incompressible materials of geo- 
physical studies, many past experi- 
ments may now have to be redone in 
part or in whole with the new media. 

As already noted, determination of 
the most abundant aqueous complexes 
in supercritical fluids is an important 
step toward providing the data neces- 
sary in modeling the transport and dep- 
osition of ore-forming minerals in the 
earth's crust. Identification of aqueous 
complexes at elevated temperatures 
and pressures with Raman spectros- 
copy is a potentially powerful method. 
Frantz has developed an optical cell for 
use to 500 ° C and 3 kbar. The disk-shaped 



cell (2" diameter, 3 k" thick) has sap- 
phire windows placed at 90° to one 
another. The system is pressurized 
with a hydraulic pump placed in series 
with a small intensifier. Preliminary 
Raman spectra on H 2 indicate that an 
exciting new tool for fluid speciation 
under geologically relevant conditions 
is almost at hand. 

A reliable thermocouple temperature 
controller for furnaces reaching 1750°C 
has been designed by Hadidiacos for 
use in experimental petrology. The con- 
troller can maintain ±1°C regulation 
in a furnace made of MoSi 2 heating 
elements. The design of this controller 
employs phase-control techniques with 
current limit protection that extends 
the lifetime of the furnace. 

The analysis of stable oxygen 
isotopes in organic matter is complex. 
The products of various decomposition 
reactions are mainly CO and C0 2 with 
differing amounts of contaminating 
gases such as methane, HC1, and nitro- 
gen. Furthermore, CO must be quanti- 
tatively converted to C0 2 for mass 
spectral analysis. Hoering and Estep 
have devised a gas chromatographic 
technique for purifying the CO before 
its conversion to C0 2 and for purifying 
the combined C0 2 for its subsequent 
analysis. Carbon monoxide is purified 
without isotopic fractionation by pass- 
ing it through a gas chromatographic 
column of molecular sieve 5A, which ef- 

fectively separates CO from other non- 
condensable gases. Purified CO can 
then be quantitatively converted to 
C0 2 by electric discharge. Very slight 
impurities in the CO, such as air, result 
in the formation of N 2 (mass 44) and 
N0 2 (mass 46). As little as 50 ppm N0 2 
in C0 2 can shift the apparent isotope 
ratio by +10%o. Carbon dioxide formed 
directly from organic oxygen and from 
the reaction of CO to C0 2 is then 
purified by preparative gas chromatog- 
raphy on a Poropak column (Waters 
Associates, Milford, Massachusetts). 

The selective adsorption properties 
of the molecular sieve silicalite have 
been shown by Hoering to be useful in 
isolating monomethyl, acyclic hydro- 
carbons from petroleums. In a search 
for a more efficient analytical method, 
Freeman discovered that separation of 
normal plus branched hydrocarbons 
from the saturated hydrocarbon frac- 
tion of petroleum could be accomplished 
in a few minutes with rc-hexane or 
cyclohexane in a chromatographic col- 
umn. Direct isolation of the normal and 
branched hydrocarbons was then ob- 
tained with an estimated 95% recovery 
after a single treatment with n-octane 
at 60 °C. Freeman believes it is reason- 
able to expect that the technique used 
in isolating the monomethyl alkanes 
may be applicable to other groups of 

Hatten S. Yoder, Jr. 


High-Pressure Experiments on FeS 

with Bearing on the Composition of 

the Earth's Core* 

H. K. Mao, G. Zou, and P. M. Bell 

The iron-sulfur system (Fe-S) has at- 
tracted attention in geophysics be- 

* Research supported in part by National 
Science Foundation grant EAR8008804 and 
Department of Energy grant DE-AS05-80ER- 

cause its densities and melting proper- 
ties are predicted to be consistent with 
those of the earth's core (Ahrens, 
1979). Determination of the phase rela- 
tions in the system Fe-S by Kullerud et 
al. {Year Book 64, 197) and by Taylor 
and Mao (1970) (as modified by Ry- 
zhenko and Kennedy, 1973; Usselman, 
1975a, b; King et al, Year Book 77, 
830-835; and Pichulo, 1978) included 
the identification of two high-pressure 



phases of troilite. Shock-wave equa- 
tion-of-state data have been obtained 
on pyrite (FeS 2 ; Simakov et al, 1974) 
and pyrrhotite (Fe 09 S; Ahrens, 1979). 
Recent static and shock-wave data on 
metallic iron (e-Fe) (Mao and Bell, 1979; 
Brown and McQueen, 1981), if com- 
bined with the data on pyrite and pyr- 
rhotite, could form the basis for mak- 
ing an estimate of the properties of the 
core (Brown and McQueen, 1981). 

Accurate data on the properties of 
the more iron-rich phase, FeS (troilite), 
are needed in order to construct a com- 
positional model relevant to the known 
density of the core. Data on the phase 
equilibria and the pressure dependency 
of the specific volume of FeS were ob- 
tained in the present study to a 
pressure of 600 kbar. The results pro- 
vide the first experimentally deter- 
mined value for the composition of the 
core consistent with earthquake data 
and shock-wave Hugoniot results. 
With additional refinement, data on 
FeS stable at 600 kbar may be useful in 
calculating melting relations and other 
properties of the core. 

Experimental Methods 

Experiments were run on pure FeS 
(99.99 wt %), synthesized by N. Boc- 
tor, in the diamond-window, high-pres- 
sure cell. Powder x-ray diffraction data 
were obtained in situ at high pressures, 
which were calibrated by the ruby fluo- 
rescence method (Mao et al, 1978). A 
work-hardened, stainless steel (T301) 
gasket was used to confine the sample, 
which consisted of powdered FeS plus 
a few grains (approximately 10 /xm in 
diameter) of ruby. No internal stan- 
dard (or pressure medium) other than 
ruby grain(s) was included in order to 
avoid possible obscuration of x-ray dif- 
fraction lines of FeS. The properties of 
the starting material are identical with 
those listed for standard FeS by the 
American Society for Testing and Ma- 
terials. The experiments were run in 
the range 30-603 kbar. 


A transition in troilite at 35 kbar, 
originally found by Pichulo (1978), was 
determined by crystallographic meth- 
ods to proceed from a distorted NiAs 
structure type [FeS(I)] to a highly 
distorted NiAs structure type 
[FeS(III)] (Fig. 1). Taylor and Mao 
(1970) found a new phase of FeS, des- 
ignated here as FeS (IV), at 130 kbar at 
ambient temperature. A similar phase 
was noted by Pichulo (1978) at the 
same pressure, and although many of 
the diffraction lines coincided in spac- 
ing with those observed by Taylor and 
Mao, several lines did not; further- 
more, the intensities of the lines that 
coincided were different. In the present 
study, this phase was observed to have 
an x-ray diffraction pattern identical 
with that of the high-pressure phase 
found in the study of Taylor and Mao 

Pichulo (1978) assigned an ortho- 
rhombic unit cell to fit his x-ray diffrac- 
tion data for the high-pressure phase 
[FeS(IV)], but his assigned cell does not 
fit the present data. The present x-ray 
diffraction data on FeS(IV) fit a dif- 
ferent orthorhombic unit cell (Table 1). 
The orthorhombic cell has a distorted 
Bl (NaCl-type) structure (Fig. 2); the S 
atoms are in cubic close packing. The 
orthorhombic lattice parameters (a, b, 
c) are related approximately to a cubic 
cell edge (a 1 ) by the function a — V2 a', 
6 = V2 a', c = 2a', for z = 16. The 
x-ray diffraction data, obtained in situ 
at high pressure, were indexed to this 
cell, and the specific volume of FeS(IV) 
was calculated as a function of pres- 

The specific volume data obtained on 
FeS(IV) from 115 to 600 kbar (Fig. 1) 
were fit to a Murnaghan equation of 
state as follows: 

V/V oh = [1 4- (K Q 'IK )P}-^o\ 

where VIV 0h is the relative change of 
the molar volume of the high-pressure 





0.7 - 

i 1 1 r 

J Present data 

a King S Prewitt, 1981 

Fig. 1. Relative volume (VI V Q ) of FeS as a function of pressure (H. King and C. Prewitt, personal 
communication, 1981). 

phase, K is the isothermal bulk mod- 
ulus, and K' is the first pressure deriv- 
ative of K. The subscript zero refers to 
kbar pressure. The parameters used 
to fit the data (Fig. 1) were V — 
18.17(3) cm 3 /mol, V oh = 15.70 ± 0.15 
cm 3 /mol, K (Mbar) = 1.37 ± 0.18, and 
K ' = 4.0. 

TABLE 1. FeS at 320 kbar and 293 K 


d (obs), A 

d (calc), A 












































































In applying these data on the Fe-S 
system to models of the composition of 
the earth's core, one notes several fac- 
tors that constrain the models. The 
densities of the high-pressure e phase 
of iron (Mao and Bell, 1979), when re- 
duced to temperature-pressure condi- 
tions of the liquid (outer) core (Masters, 
1979), are close to seismically deter- 
mined values of density (Gilbert and 
Masters, 1980) but actually fall outside 
the limits of uncertainty. Brown and 
McQueen (1981) confirmed that pure 
iron is denser than the core and that 
additional, lighter components, such as 
sulfur, must be dissolved in core iron. 
There are other supporting arguments 
for suggesting that sulfur is a compo- 
nent of the core. Murthy (1976) and 
Ringwood (1977) have postulated a 
"sulfur budget" for the earth based on 
abundances in the solar system. Other 
light elements may also be considered, 
but sulfur meets the geochemical and 
physical requirements of most current 
models of the core. 

Epsilon iron, by itself, is too dense 
for the core, and pure sulfur is too 



Fig. 2. Relation of the orthorhombic unit cell 
to the Bl cubic unit cell. Sulfur atoms are shown 
as shaded spheres in cubic closest packing. 

light, so a mixture in the Fe-S system 
is required. There are great differences 
in the relative compactness of the 
structures at 1 bar so that estimates of 
the appropriate mixtures are depen- 
dent on the structures of the phases in 
the Fe-S system. The densities at high 
pressure, however, are more closely 

A plot of the present data on FeS and 
the data on e-Fe (Mao and Bell, 1979) in 
terms of mean atomic volume vs. 
atomic % sulfur (in the Fe-S system) re- 
veals a linear relationship at pressures 
above 600 kbar that can be used to 
estimate the composition as a function 
of density in the Fe-S system. As the 
pressure increases to 600 kbar, the 
volume converges to dense, close- 
packed structures of all iron-bearing 
phases in the Fe-S system, shown in 
Fig. 3. 

The plotted curves of pressure vs. 
density for the solid sulfides, the liquid 
core, and e-iron are related (Fig. 4). At 
pressures above approximately 600 
kbar, the pressure vs. density curves 
are similar, evidently resulting from 
compressibility functions of the same 
general form, even though structurally 

different. The range of pressure is 
great enough so that experimental 
differences (i.e., Hugoniot, static 
pressure at ambient temperature, and 
seismological techniques) that exist in 
the combined data do not obscure the 
relationship. In simplest terms, the 
most important factor affecting the 
pressure-density functions of Fe-S sol- 
ids and liquids at sufficient pressure is 
the composition along the join. 

The results can be used to calculate 
the composition of the core in terms of 
iron plus "light element" (subject, of 
course, to assumptions as to the light 
element or elements that exist). Calcu- 
lations, including thermal corrections, 
of the present static data and the ex- 
isting Hugoniot data indicate that sul- 
fur by itself, as a light element compo- 
nent in models of an otherwise pure 
iron core, is acceptable (Ahrens, 1979) 
(Fig. 4). To determine the percentage of 
sulfur from the present data, melting 
values of pure iron and of the eutectic 
composition of the Fe-S system are re- 
quired at core pressures. The melting 
curves, although uncertain (±500°C), 


1 — 

Moo ond Bell, 1979 
Present data 
Ahrens, 1979 
Simakov et_al., 1974 

Fig. 3. Plot of mean atomic volume vs. 
atomic percent sulfur in the Fe-S system. 




of 2 

FeS 2 23°C isotherm from 
shock-wave data of 
Simakov et al„ 1974 

Fe 09 S 23°C isotherm 
from shock-wave data 
of Ahrens, 1979 

FeS King S 
Prewitt, 198 

/ vW 


€-Fe 23°C isotherm from 
shock-wave data of 
Brown 8 McQueen, 1981 

5 10 15 

Density, gm/cm 3 

Fig. 4. Pressure-density plot. (The general form of the diagram is after Ahrens, 1979; data of 
King and Prewitt are from personal communication, 1981.) 

are of sufficiently flat trajectory 
(Brown and McQueen, 1981) that anal- 
ysis of their intersection with the equa- 
tions of state can be used to yield com- 
position values. 

The intersection of the curves for the 
sulfide solids with the seismologically 
derived data for the core and with the 
equation of state of e-Fe yields a value 
of sulfur of approximately 7-9 wt % 
(Fig. 4). This value is in agreement with 
that derived by Ahrens (1979) from an 
analysis of sulfide shock-wave data 
alone and is within the range of sulfur 
composition deduced by Brown and 

McQueen (1981) from shock-wave and 
static equations of state for e-Fe. 

A comparison of this model value for 
sulfur content in the outer core with 
the so-called sulfur budget for the 
earth indicates that the value of 7-9 wt 
% sulfur for the core is lower than val- 
ues recently estimated for the whole 
earth (Jeanloz, 1981). Thus, if solar 
system abundances apply to the 
earth's composition, either sulfur is rel- 
atively depleted in the earth or it must 
be a component of the mantle as well as 
the core to compensate for reduced 
sulfur in the core. 




Ahrens, T. J., Equations of state of iron sulfide 
and constraints on the sulfur content of the 
earth, J. Geophys. Res., 84, 985-998, 1979. 

Al'tshuler, L. V., K. K. Krupnikov, B. N. Led- 
enev, V. I. Zhuchikhin, and M. I. Brazhnik, 
Dynamic compressibility and equation of 
state of iron at high pressures, Zh. Eksp. Teor. 
Fiz., 34, 4, 1958. 

Brown, J. M., and R. B. McQueen, The equation 
of state of iron and the earth's core, in High- 
Pressure Research: Applications in Geophys- 
ics, S. Akimoto and M. Manghani, eds., Center 
for Academic Publications of Japan, Tokyo, in 
press, 1981. 

Gilbert, J. F., and G. Masters, Stacking for mul- 
tiplet periods (abstract), Eos, 61, 1043, 1980. 

Hart, R. S., D. L. Anderson, and H. Kanamori, 
The effect of attenuation on gross earth mod- 
els, J. Geophys. Res., 82, 1647-1654, 1977. 

Jeanloz, R., High-pressure chemistry of the 
earth's deep interior, in American Chemical 
Society 17th State-of-the-Art Symposium, 
Washington, D.C., in press, 1981. 

Mao, H. K., and P. M. Bell, Equation of state of 
MgO and e-Fe under static pressure condi- 
tions, J. Geophys. Res., 84, 4533-4536, 1979. 

Mao, H. K., P. M. Bell, J. W. Shaner, and D. J. 
Steinberg, Specific volume measurements of 
Cu, Mo, Pd, and Ag and calibration of the 
ruby R l fluorescence pressure gauge from 0.06 
to 1 Mbar, J. Appl. Phys., 49, 3276-3283, 1978. 

Masters, G., Observational constraints on the 
chemical and thermal structure of the earth's 
deep interior, Eos, 57, 507-534, 1979. 

Murthy, V. R., Composition of the core and the 
early chemical history of the earth, in The 
Early History of the Earth, B. F. Windley, ed., 
John Wiley and Sons, New York, pp. 21-31, 

Pichulo, R. O., Polymorphism and phase rela- 
tions in FeS, Ph.D. thesis, Columbia Univer- 
sity, New York, 1978. 

Ringwood, A. E., Composition of the core and 
implications for the origin of the earth, Geo- 
chem. J., 11, 111-136, 1977. 

Ryzhenko, B., and G. C. Kennedy, The effect of 
pressure on the eutectic in the system Fe-FeS, 
Amer. J. Sci., 273, 803-810, 1973. 

Simakov, G. V., M. N. Paulovskiy, N. G. Kala- 
shnikov, and R. F. Trunin, Shock compress- 
ibility of twelve minerals, Izv. Acad. Sci. 
USSR, Phys. Solid Earth, 8, 11-17, 1974. 

Taylor, L. A., and H. K. Mao, A high-pressure 
polymorph of troilite, FeS, Science, 170, 850- 
851, 1970. 

Usselman, T. M., Experimental approach to the 
state of the core, Part 1, Amer. J. Sci., 275, 
278-290, 1975a. 

Usselman, T. M., Experimental approach to the 
state of the core, Part 2, Amer. J. Sci., 275, 
291-303, 1975b. 

Application of the Solid-Helium 

Pressure Medium in a Study of the 

a-e Fe Transition Under 

Hydrostatic Pressure* 

G. Zou, P. M. Bell, and H. K. Mao 

The various thermodynamic states 
of metallic iron could have important 
bearing on the formulation of geophys- 
ical models of the earth's core and core- 
mantle boundary. It is not yet known, 
however, which phase of iron is stable 
in the core (Brown and McQueen, 
1980). The phase relations between the 
a (body-centered cubic), 7 (face-cen- 
tered cubic), and e (hexagonal close- 
packed) forms of solid iron are known 
only approximately from the results of 
nonhydrostatic experiments (Mao et 
al, 1967; Strong and Bundy, 1971). 
Nonhydrostatic compressibility data 
were obtained to 1 Mbar in a recent 
study {Year Book 76, 519-522). The 
present investigation of the a-e transi- 
tion and the compressibility of iron 
near the phase boundary under hydro- 
static conditions is the first such study 
done with solid helium as the pressure- 
transmitting medium. 

In addition to geophysical applica- 
tions, the a-e iron transition under am- 
bient conditions has been used as a 
fixed-point pressure calibration (Strong 
and Bundy, 1971). Detailed data on the 
transition in solid-media, high-pressure 
apparatus indicate that even though the 
reaction from a to e and its reverse are 
rapid, a two-phase "hysteresis" zone 
can exist over a 100-kbar range (Mao et 
al, 1967; Giles et al, 1971). 

The present data were obtained in 
two independent sets of experiments in 
which helium surrounded samples of 
pure iron metal (99.95 wt % Fe) held 
under pressure at ambient temperature 
in the diamond-window, high-pressure 
cell. In one set, the sample was finely 

* Research supported in part by National 
Science Foundation grant EAR8008804 and 
Department of Energy grant DE-AS05-80ER- 



divided powder pressed into the form 
of a plate; in the second set, the sample 
was in the form of a folded section of 
thin (10-/xm-thick) foil or sheet. Detec- 
tion of the a (bcc) and e (hep) phases 
was accomplished by x-ray diffraction. 
A rotating- anode, Mo-target, x-ray 
source was coupled to a quartz-source 
monochrometer to provide MoK a] 
radiation for the diffractometer mea- 

Pressure was measured by the ruby 
fluorescence method. The hydrostatic 
behavior of the pressure medium was 
evaluated by measuring the pressure in 
several parts of the sample chamber 
and by measuring the resolution of the 
ruby Ri fluorescent line. The former 
was done to test for pressure gradients 
or pressure inhomogeneity; the latter, 
for stress difference in ruby. In both 
sets of experiments the pressure gradi- 
ents and the stress differences were 
minimal or nonexistent, and yet there 
was apparent sluggishness of the 

In the first set of experiments (Table 
2), complete reaction from the a phase 
to the e phase was observed at approxi- 

TABLE 2. In Situ High-Pressure X-Ray 
Diffraction Data on the a(bcc)-( (hep) Phase 
Transition in Fe Metal (Powdered Sample) 

mately 152.7 kbar on the increasing 
pressure cycle. (Very weak diffraction 
lines that might be attributed to the e 
phase were observed in the data at 
137.1 kbar.) In the second set (Table 3), 
the e phase was observed at approx- 
imately 154.7 kbar on the increasing 
cycle. These values contrasted sharply 
with the range of previously published 
values of the a-e Fe "fixed point" of 
approximately 107-130 kbar, and thus 
the transition must be used with cau- 
tion as a pressure reference. The 
explanation of the differences ap- 
parently lies in the degree of uniform 
nonhydrostatic stress of the different 
experiments. In pressure-increase cy- 
cles of these experiments, the helium 
pressure medium was sufficiently hy- 
drostatic to avoid formation of a two- 
phase region. In former studies, a two- 
phase region at ambient temperature 
could have been stabilized by nonhy- 
drostatic stress to a pressure below the 
hydrostatic value, whereas in the 
present experiments the low-pressure 
phase (a) was held metastably to pres- 
sures above its stability field, probably 
because the transition was sluggish in 
the absence of shearing stress. 

The pressure of the a-e iron transi- 
tion obtained on the decompression cy- 
cles of the present experiments is less 

Pressure, kbar 

Observed Phases 












a(bcc) + tr(?) 



Decompression cycle: 

144.0 e(hcp) 

124.7 e(hcp) 

123.8 e(hcp) 
109.6 ((hep) 

92.0 a(bcc) + ((hep) 

88.5 a(bcc) + ((hep) 

83.5 a(bcc) + t(hcp) 

80.0 a(bcc) + ((hep) 

67.4 a(bcc) 

TABLE 3. In Situ High-Pressure X-Ray 

Diffraction Data on the a(bcc)-e(hcp) Phase 

Transition in Fe Metal (Foil Sample) 

Pressure, kbar Observed Phases 











Decompression cycle: 






((hep) + a(bcc) 


((hep) + a(bee) 


((hep) + a(bcc) 


((hep) + a(bcc) 


a (bcc) 



certain, but the data are listed in 
Tables 2 and 3 for comparison. The un- 
certainty arises because on the decom- 
pression cycle the major response of 
the stressed parts of the high-pressure 
cell is elastic (as opposed to flow of the 
gasket on compression), and thus, with 
less displacement, the pressure falls 
very rapidly on unloading. Slight leaks 
of helium have been observed on the 
decompression cycle with resultant 
readjustment of the gasket. It is likely 
that nonhydrostatic (and possibly 
residual) stresses exist in the sample 
(although not in the helium) on decom- 
pression; if so, these stresses would 
help to explain the two-phase 
hysteresis observed (Tables 2 and 3). 
The two-phase region was thus 
stabilized to pressures as low as 80 
kbar (Table 2). The low-pressure phase 
of iron (a) was not observed in either 
set of experiments until the pressure 
was released below approximately 109 
kbar, a result indicating metastability 
of the high-pressure phase (e). The suc- 
cessful use of solid helium as a pressure 
medium has, however, revealed the 
true nature of the a-e iron transition as 
one characterized by a sluggish reac- 
tion rate at ambient temperature. 
Confirmation of the absence of two- 
phase regions plus the hysteresis (over- 
shoot) is a demonstration of the slug- 
gishness of the reaction. 


Brown, J. M., and R. G. McQueen, Melting of 
iron under core conditions, Geophys. Res. 
Lett, 7, 533-536, 1980. 

Giles, P. M., M. H. Longenbach, and A. R. Mar- 
der, High-pressure martensitic transforma- 
tion in iron, J. Appl. Phvs., 42, 4290-4295, 

Mao, H. K., W. A. Bassett, and T. Takahashi, 
Effect of pressure on crystal structure and lat- 
tice parameters of iron up to 300 kbar, J. Appl. 
Phys., 38, 272-276, 1967. 

Strong, H. M., and F. P. Bundy, The use of iron 
and gold for calibration of higher pressure and 
temperature points, U.S. Nat. Bur. Stand. 
Spec. Publ, 326, 283-290, 1971. 

Irreversible Unit-Cell Volume 

Changes of Wustite Single 

Crystals Quenched from High 


R. M. Hazen, H. K. Mao, L. W. Finger, 
and P. M. Bell 

An underlying assumption of 
hydrostatic compression experiments 
is that unit-cell volumes adjust rapidly 
and reversibly to changes in pressure 
and temperature. This ideal elastic be- 
havior has been confirmed in hundreds 
of elements and compounds, including 
all minerals studied before the present 
experiments. Single-crystal wustite 
(Fe .947O), on the other hand, has been 
found to undergo what appears to be a 
gradual irreversible reduction in unit- 
cell volume when subjected to hydro- 
static pressures above 150 kbar. High- 
precision, single-crystal, x-ray diffrac- 
tion techniques were used to detect 
this phenomenon. 

Magnesiowiistite has been proposed 
as a major iron-bearing phase in the 
earth's lower mantle (Mao and Bell, 
1977). The P-V equation of state of 
wustite (Fe!_. r O, 0.85 < jc < 1.00) is 
complicated by nonstoichiometry and 
a possible reversible, second-order 
phase transition to a distorted form 
over the pressure range 50-180 kbar 
(Year Book 79, 374-380). The present 
study of the hydrostatic compression 
of single-crystal wustite was originally 
undertaken to document volume and 
symmetry changes near the reversible 
transition. In hydrostatic pressure 
experiments, as compared with the 
nonhydrostatic powder experiments of 
the 1980 report, no deviations from cu- 
bic symmetry were detected up to 150 
kbar. [Transition pressure may, how- 
ever, be affected by the absence of 
shear (see Zou, Bell, and Mao, this 
Report).] The present experiments pro- 
vide information on the equation of 
state of wustite, as well as on its 

♦Research supported in part by National 
Science Foundation grant EAR79 19768. 



previously unreported irreversible be- 

Single crystals of synthetic wiistite 
of the approximate composition Feo.9470 
were examined with a four-circle auto- 
mated diffractometer. Unit-cell dimen- 
sions were determined from positions 
of wiistite diffraction corrected for 
systematic errors in crystal centering 
and diffractometer alignment (King 
and Finger, 1979). Unit-cell volumes of 
nine crystals, ranging in size from 40 to 
120 ixm maximum dimension, were 
determined before pressure treatment. 
Volumes ranged from 79.957(8) to 
79.980(8)A 3 , with an average value of 
79.97(1) A 3 . No systematic variation in 
unit-cell volume with crystal size was 
observed, and it is thus assumed that 
systematic errors due to specimen ab- 
sorption are not significant. 

A crystal plate, 100 X 100 X 40 /mi, 
was mounted in a triangular diamond 
pressure cell with methanol-ethanol 
pressure fluid as described by Hazen 
and Finger (Year Book 76, 655-656). 
The unit-cell volume at approximately 
50 kbar was measured, and no devia- 
tions from cubic symmetry were de- 
tected; the three cell edges ranged from 
4.264(14) to 4.274(9) A, and angles 
deviated by no more than 0.1° ± 0.2° 
from 90°. After these high-pressure ex- 
periments, which lasted 7 days, the 
crystal had returned to a volume of 
79.97 A 3 . 

A second crystal, of maximum 
dimension 60 jum, was mounted in a 
cryogenically loaded cell {Year Book 
79, 409-411). In these experiments the 
crystal of FeO is located within a weak 
single crystal of neon, which acts as a 
hydrostatic medium. (This pressure 
cell is capable of hydrostatic exper- 
iments on single crystals to pressures 
in excess of 200 kbar.) Unit-cell 
volumes of wiistite at several pressures 
from 85 to 144 kbar were determined 
(see Table 4), and again no distortions 
from cubic symmetry were detected. 
Higher pressures were attempted, but 
this mount relaxed from 160 kbar to 

TABLE 4. Unit-Cell Volume of Wiistite 
(Feo.947 0) vs. Pressure 











Average of 
10 crystals 























































less than 150 kbar. After 21 days of 
pressures between 120 and 150 kbar, 
the unit-cell volume of crystal No. 2 
was 79.74(2) A 3 (see Table 5). This 
volume is 0.3%, or more than ten stan- 
dard deviations smaller than the initial 

A third crystal, with maximum di- 
mension 50 fim, was selected and pres- 
surized with neon to 167 kbar, at which 
point another unit-cell volume mea- 
surement was made. At higher pres- 
sure this sample was crushed between 
the diamonds. 

A fourth, smaller crystal was pres- 
surized to 170 kbar for 14 days and 
then 200 kbar for 14 days. Unfortu- 
nately, the 30-/xm crystal was too small 
to produce measurable diffraction 
through the diamonds of the pressure 
cell. After 28 days the cell was opened, 
and the volume of the small crystal 
was redetermined under room condi- 
tions. The resultant unit-cell volume of 
79.34 A 3 was 0.8% smaller than the 
original value. 

As an additional test of this irrever- 
sible behavior, two 100-/xm equant 



TABLE 5. Unit-Cell Volume of Wustite (Fe . 9 47O) Before and After Hydrostatic 

Pressure Treatment 

Crystal No. 

Pressure History 

Volume, A 3 



Before pressure 




Before pressure 



Average of 

10 crystals 

Before pressure 




3 days at 215 kbar in neon 




3 days at 215 kbar in neon 




21 days at 120-150 kbar in 





28 days at 170-200 kbar in 




crystals with cell volumes before pres- 
surization of 79.974(6) and 79.960(8) A 3 
were broken, and fragments of less 
than 50 /mi maximum diameter were 
subjected for 3 days to 215 kbar, which 
is the highest pressure yet attained in 
any single-crystal experiment. Upon 
pressure release, the cell volumes were 
found to be 79.76(3) and 79.73(5) A 3 , 
corresponding to volume reductions of 

There is strong evidence for a time- 
and pressure-dependent irreversible 
volume change in wustite. All nine 
crystals measured before pressuriza- 
tion had volumes of 79.97(1) A 3 , 
whereas all samples subjected to 
pressures greater than 150 kbar 
showed volumes of 79.76 A 3 or less. 
Furthermore, the sample that was held 
at 170-200 kbar for 28 days showed a 
significantly greater reduction than 
samples held at 215 kbar for 3 days or 
120-150 kbar for 21 days. This type of 
time and pressure dependence is sim- 
ilar to the time-temperature-transfor- 
mation behavior associated with diffu- 
sion processes. 

The nature of the irreversible wustite 
compression is not yet clear. At high 
temperatures of synthesis, pressure 
has a significant effect on wustite 
stoichiometry. Above 100 kbar, for ex- 
ample, the iron content of synthetic 
wustite decreases with pressure 
(Simons, Year Book 79, 376-380) as a 
result of the greater density of iron 
metal plus Fe 3+ -rich wustite compared 

with pure FeO. It is possible that ex- 
solution of metallic iron at the unit-cell 
level might cause the observed irrever- 
sible phenomena, though such com- 
paratively rapid exsolution at room 
temperature would be surprising. Ex- 
solution of iron would be both time and 
pressure dependent. Reordering of de- 
fects and ferric iron at high pressure 
might also explain the observed vol- 
ume decreases without requiring a 
compositional change in the wustite 
phase. Details of such an ordering of 
vacancies or Fe 3 + cations are not ob- 
vious. Irreversible changes have been 
observed only in experiments employ- 
ing neon as the hydrostatic pressure 
medium. The possibility of time-depen- 
dent interaction of neon with the non- 
stoichiometric sample cannot yet be 
ruled out, although the introduction of 
neon into the wustite structure should 
not result in a volume decrease. 

Pressure-volume data in Table 4 may 
be used to calculate equation-of -state 
parameters for wustite (Fig. 5). It must 
be noted that any irreversible volume 
changes induced at high pressure will 
tend to lower the calculated bulk mod- 
ulus (K) and perhaps reduce the ob- 
served curvature (K') as well. If the 
total irreversible change is only 0.3% 
(corresponding to a pressure of about 5 
kbar), then the effect on K will be no 
greater than 3%, which is approx- 
imately twice the estimated error of 
the present experimental data to 167 
kbar. Values of K and K' calculated 




1 ' 

1 ' 

1 ' 1 




(Fe 0947 0) 
K= 1.54(2) 
K' = 2.1 (4) 




v °- 94 

v o 


N \ 






\^ " 





50 100 

Pressure, kbor 

Fig. 5. Compressibility of single-crystal wustite Fe 947 0. 

from a second-order Birch-Murnaghan 
equation of state are K = 1.54(2) Mbar 
and K' = 2.1(4). Note that equations of 
state based on compression data to 
pressures significantly above 200 kbar 
may be severely affected by irrevers- 
ible effects. Special care must be taken 
to measure the sample both before and 
after compression to document these 

The advent of the cryogenically 
loaded cell has provided a new range of 
pressure conditions for single-crystal 
experiments. Neon and argon have 
both been used to pressures above 100 
kbar, and helium and hydrogen may be 
used as hydrostatic media to pressures 
above 600 kbar (this Report). As crys- 
tals are subjected to these extreme con- 
ditions it may be necessary to reex- 
amine conventional assumptions about 
the effects of pressure on the crystal- 
line state. 


King, H. E., and L. W. Finger, Diffracted beam 
crystal centering and its application to high- 
pressure crystallography, J. Appl. Cryst, 12, 
374-378, 1979. 

Mao, H. K., and P. M. Bell, Disproportionation 

equilibrium in iron-bearing systems at pres- 
sures above 100 kbar with applications to 
chemistry of the earth's mantle, in Energetics 
of Geological Processes, S. K. Saxena and S. 
Bhattacharji, eds., Springer-Verlag, New 
York, pp. 236-249, 1977. 

Systematic Variation of Bulk 

Modulus of Wustite with 


R. M. Hazen 

Measurement of small differences in 
bulk moduli of two compounds is a dif- 
ficult experimental problem. Individ- 
ual bulk moduli of oxides and silicates 
may have reported precisions of ±1%, 
yet systematic instrumental and pres- 
sure-calibration errors commonly lead 
to discrepancies of ±10% between pub- 
lished values. A description of a new 
procedure that facilitates the measure- 
ment of differential, as well as absolute, 
compression of two or more single crys- 
tals is reported herein. It is possible 
with this procedure to detect differ- 
ences of less than 1 % in the bulk moduli 
of two oxide or silicate samples. 

* Research supported in part by National 
Science Foundation grant EAR79 19768. 



The new technique has been used to 
study the problem of the effect of non- 
stoichiometry on the compressibility of 
wiistite (Fe^.,.0). Several authors have 
measured the bulk moduli of wiistite 
(Mao et al., 1969; Jackson et al, 1978; 
Jeanloz and Ahrens, 1980; Hazen et al, 
this Report), and a range of values from 
1.45 to 1.85 Mbar has been reported. In 
general, static compression experi- 
ments have yielded significantly smaller 
bulk moduli of wiistite than dynamic 
compression techniques, but no sys- 
tematic trends with composition have 
been conclusively demonstrated. 

Differential and absolute compres- 
sion of three crystals of wiistite were 
measured by placing 0.04-0.06 mm di- 
ameter samples of Fe .9oO, Fe .93O, and 
Fe 947O (unit-cell edges, respectively, 
a = 4.2903, 4.3017, and 4.3083, all 
±0.0001 A) together in the sample 
chamber of a diamond-window, high- 
pressure cell. A major advantage of us- 
ing multiple crystals in one mount is 
that the hydrostatic pressure is iden- 
tical for all crystals, so that the largest 
source of uncertainty is eliminated. 

Crystals were attached to one diamond 
face with silicone vacuum grease, and 
the crystals, along with ruby calibra- 
tion chips, were arranged to minimize 
x-ray interference between crystals and 
the gasket (see Fig. 6). Methanol- 
ethanol (4:1) was used as the hydrostatic 
pressure medium in a single-crystal 
pressure cell for x-ray diffraction stud- 
ies (Hazen and Finger, Year Book 76, 

Reflections of class [400] (29 = 39°, 
MoK ai ) were measured for each wiis- 
tite crystal with an automated four-cir- 
cle diffractometer. The 8-reflection cen- 
tering procedure (King and Finger, 
1979), which eliminates crystal center- 
ing and diffractometer zero errors, was 
used. All (400) reflections were mea- 
sured at least six times in a 20-hour 
period, and pressure was calibrated 
both before and after each set of 
20 measurements. Reflections for the 
Fe 090 O and Fe 947O samples were re- 
producible to ±0.001° 29; the (400) re- 
flection of Fe 93 0, which was a smaller 
crystal, was reproducible to ±0.003°. 
The resulting fractional uncertainty in 

Fig. 6. Multiple crystal mount for high-pressure studies of the relative and absolute compression 
of wiistite. (A) View in transmitted light; (B) cut-away side view. Black crystals are three different 
wustite compositions; colorless crystals are chips of ruby for pressure calibration. The crystals were 
arranged to minimize x-ray interference between crystals and the gasket. The relative positions and 
orientations of the three wustite crystals remained close to this configuration throughout the ex- 



volume is approximately 0.0001 for 
Fe 090 O and Feo 947 0, and 0.0003 for 
Fe . 93 O. 

Compression of the three wiistite 
crystals at several pressures to 54.7 
kbar can be compared directly (Table 
6). At all pressures the compression of 
Fe 90 O is greater than or equal to that 
of Fe .947O. The average compression 
of Fe .93O is intermediate to that of the 
other two crystals. Figure 7 is a graph 
of V/V vs. pressure for both Fe 90 and 
Fe 094 7 wiistites. At the bottom of the 
graph is plotted the difference between 
V/Vq for these two crystals vs. pres- 
sure. Linear regression of these data on 
difference vs. pressure indicates that 
Fe 09 oO is 1.4 ± 0.4% more compressi- 
ble than Fe 09 47O. Bulk moduli of 
Fe 09 oO, Fe 093 O, and Fe 0947 O, with K' 
= 2.1 (see Hazen et al, this Report), are 
1.519(6), 1.531(8), and 1.540(7) Mbar. 
Note that the absolute uncertainties in 
bulk moduli are at least ±5% because 
of uncertainties in K'. The relative un- 
certainty of the three bulk moduli, how- 
ever, is less than 0.5%, and a small but 
significant increase in compressibility 
with degree of nonstoichiometry is evi- 

A major unresolved question regard- 
ing the equation of state of wiistite 

Fig. 7. Absolute (A) and relative (B) compres- 
sion of two crystals of wiistite, Fe 090 O and 
Fe .947O- 

is the discrepancies between static 
and dynamic compression studies. All 
static compression measurements are 
consistent with the 1.5-Mbar bulk mod- 


Compression of Single-Crystal Wiistite in 

a Multiple 

Crystal Mount 

VI V * 


x = 0.90 

x = 0.93 

x = 0.947 






(in air) 





9 (in cell) 








































* Unit-cell volumes of Fe .9oO, Fe .93O, and Fe .947O are 78.969. 79.601, and 
79.971, all ±0.006 A 3 , based on single-crystal x-ray diffraction of wiistite in 



ulus found in this study; bulk moduli 
greater than 1.6 Mbar are precluded by 
these static data. All dynamic com- 
pression studies are consistent with a 
value of 1.8 Mbar; bulk moduli smaller 
than 1.7 Mbar are inconsistent with dy- 
namic compression results (Jeanloz, 
personal communication, 1981). Possi- 
ble sources of systematic differences, 
including sample preparation and ef- 
fects of bulk viscosity, are now being 


Jackson, I., R. C. Liebermann, and A. E. Ring- 
wood, The elastic properties of Mg r Fe 1 _ J .0 
solid solutions, Phys. Chem. Minerals, 3, 
11-31, 1978. 

Jeanloz, R., and T. J. Ahrens, Equations of state 
of FeO and CaO, Geophys. J. Roy. Astron. 
Soc, 62, 505-528, 1980. 

King, H. E., and L. W. Finger, Diffracted beam 
crystal centering and its application to high- 
pressure crystallography, J. Appl. Cryst., 12, 
374-378, 1979. 

Mao, H. K., T. Takahashi, W. A. Bassett, J. S. 
Weaver, and S. Akimoto, Effect of pressure 
and temperature on the molar volumes of 
wiistite and of three (Fe,Mg) 2 Si0 4 spinel solid 
solutions, J. Geophys. Res., 74, 1061-1069, 

Equations of State for Thorium 

Metal, U0 2 , and a High-Pressure 

Phase of U0 2 to 650 kbar* 

T. M. Benjamin,^ G. Zou, H. K. Mao, and 
P. M. Bell 

Two cubic actinide materials, tho- 
rium metal and U0 2 , were chosen for 
the initial phase of a more comprehen- 
sive high-pressure study of actinide 
compounds. The ambient temperature 
and compressibility of these two mate- 
rials were determined to characterize 
their densities at high pressure. 

* Research supported in part by National 
Science Foundation grant EAR8008804 and 
Department of Energy grant DE-AS05- 

fCurrent address: Los Alamos National Labo- 
ratory, CNC-7, MS-514, Los Alamos, New Mex- 
ico 87545. 

Thorium metal had been predicted by 
Liu (1973) to undergo a phase transfor- 
mation on the basis of his correlation 
between density and previously ob- 
served, pressure-induced phase trans- 
formations in similar metals. Previous 
study at static pressures below 100 
kbar produced no phase transforma- 

The current experiments were con- 
ducted in the diamond-window, high- 
pressure cell to 677 kbar, as measured 
by the ruby fluorescence technique 
(Mao and Bell, Year Book 77, 910), on 
ground-in-alcohol thorium metal (Alfa 
Chemicals). X-ray diffraction data 
(Table 7) were recorded on film from 
MoK a radiation and a Pd-metal exter- 
nal standard. The unit-cell parameter 
of thorium metal measured at room 
pressure was 5.0743 ± 0.0096 A. 

The equation-of-state (EOS) param- 
eters were determined by the method of 
weighted nonlinear least-squares. Only 
the errors in pressure contributed to 
the weighting, but the agreement be- 
tween the V/Vq data and the errors can 
be used to judge the adequacy of the fit 
(Fig. 8). It should be noted that no 
phase transformation was observed to 
677 kbar. The pressure errors were in- 
creased until a reduced chi-squared less 
than two was achieved. The constant 
percentage error or the observed error 
in pressure measurement was used in 
each calculation, whichever error was 
the larger. The resultant EOS param- 
eters are K = 0.72 ± 0.02 Mbar and 
K ' = 2.0 ±0.1 for the Murnaghan 
EOS, and K = 0.67 ± 0.02 Mbar and 
K ' = 3.0 ± 0.1 for the Birch-Murnag- 
han EOS. These values are in agree- 
ment with the data derived from shock 
waves at the Los Alamos National 
Laboratory (Marsh, 1980, 139) of K = 
0.633 Mbar and K ' = 4. 

The Murnaghan EOS fits the data 
slightly better than the Birch-Mur- 
naghan EOS in that the former met the 
criteria for an acceptable fit with a 
pressure error of only 3% whereas the 
latter resulted in a pressure error of 



TABLE 7. X-Ray Diffraction Data on Th and U0 2 at High Pressures 


a, A 

b, A 

c, A 


(cm 3 /mol) 

VIV % 

Thorium Metal* 



















Cubic U0 2 , Z = 4 

0.001 5.470(8) 

65(4) 5.418(7) 

125(1) 5.377(4) 

214(4) 5.332(9) 

254f 5.321(10) 

267(2) 5.318(7) 

332(7) 5.269(7) 

Orthorhombic UO2, Z = 4 

278(4) 3.51(1) 5.90(2) 

294(5) 3.50(2) 5.91(3) 

308(5) 3.58(4) 5.93(7) 

402(3) 3.45(4) 5.81(7) 

402(4) 3.50(4) 5.80(7) 

468(3) 3.45(2) 5.79(3) 

495(5) 3.48(4) 5.79(5) 

600(10) 3.46(5) 5.77(5) 




















































* All errors are la standard deviations. 

fOnly one set of pressure data; all others are averages of data sets taken before and after x-ray 

{All pressure data corrected for the parabolic pressure distribution over the region of the sample 
exposed to x rays. 

§ V = 24.63 cm3/mol for U0 2 . 

Pressure, kbor 


Fig. 8. Normalized volume data on Th to 678 kbar. 



4%. It appears, however, that a better 
fit to the data could be obtained with a 
higher order EOS to provide more cur- 
vature at higher pressure. 

Gust (1981) at Lawrence Livermore 
National Laboratory concluded from 
shock-wave experiments that U0 2 un- 
derwent a phase change, probably to 
an orthorhombic PbCl 2 structure, at 
about 540 kbar on the Hugoniot. Ac- 
cording to Gust (personal communica- 
tion), the actual pressure of the U0 2 
phase transformation may be less than 
440 kbar owing to the strength of U0 2 
as deduced from elasticity theory. 

The current experiments (Table 7) 
were conducted to 640 kbar on ground- 
in-alcohol UO 2 .03 containing 99.76% 
238 U. All procedures are identical with 
those used in the thorium-metal experi- 
ments. The unit-cell parameter of U0 2 
measured at room pressure (a = 
5.4695 ± 0.0083 A) is in excellent agree- 
ment with the ASTM tables (a = 
5.4682 A). 

The EOS parameters, as determined 
by pressure-weighted nonlinear least- 
squares, are equally well fit by both the 
Murnaghan and Birch-Murnaghan sec- 
ond-order EOS although the Murnag- 
han EOS parameter has slightly smaller 
errors (Fig. 9). The calculated EOS 
parameters, K = 2.1 ± 0.1 Mbar and 

Kq = 7 ± 2, agree with values obtained 
by Hazen and Finger {Year Book 78, 
633). The Hazen and Finger bulk modu- 
lus {K = 2.30 ± 0.08) may be the bet- 
ter value as their technique (to 50 kbar) 
was ideal for its determination. This 
study, extended to much greater pres- 
sure, should yield a superior value of 
Kq (as is clear from the 100% error as- 
signed to Kq in the Hazen and Finger 
study). The difference between the K 
values obtained in this study and in the 
better K value of Hazen and Finger 
may indicate the need for a higher order 
EOS, as also suggested by the thorium- 
metal data. 

As predicted by Gust (1981), a phase 
transformation was observed in U0 2 to 
a PbCl 2 orthorhombic structure. The 
transformation occurs between 332 and 
403 kbar. Further work is under way to 
narrow this interval. The present data 
are insufficient to obtain precise unit- 
cell dimensions, much less to obtain re- 
liable EOS parameters (Table 7, Fig. 9). 
The unit-cell dimensions given in Table 
7 were calculated by (l/<i^ A /)-weighted 
linear least-squares of manually in- 
dexed x-ray lines with the PbCl 2 struc- 
ture as a reference base. Because of the 
imprecision of the unit-cell parameters, 
only an average value is given in Fig. 9. 
Although of preliminary nature, it is 

Orthorhombic U0 2 

200 400 

Pressure, kbar 

Fig. 9. Normalized volume data on U0 2 to 650 kbar. 



clear that a phase transformation from 
a cubic to an orthorhombic structure 
has occurred and that the high-pres- 
sure orthorhombic phase is approx- 
imately 5% denser than the cubic struc- 

This phase transformation may have 
significance in the analysis of theoreti- 
cal models of the earth's mantle (De- 
Paolo, 1981). The observed phase trans- 
formation, at room temperature, occurs 
at pressures well within the pressure 
range of the lower mantle (DePaolo, 
1981); however, the sign of dP/dT for 
the phase transformation could be neg- 
ative (a result that is not without prece- 
dent). Further experimentation on 
U0 2 and other actinide (and potassium) 
compounds is required to judge their 
role in the mantle. 


DePaolo, D. J., Nd isotopic studies: some new 
perspectives on earth structure and evolution, 
Eos, 62, 137-140, 1981. 

Gust, W. H., Elastic and plastic properties of 
uranium dioxide from 5 to 330 GPa, Proc. 7th 
AIRAPT Int. High Press. Conf., France, 1979, 
in press, 1981. 

Liu, L.-G., Correlation of density periodicity to 
pressure-induced polymorphic transforma- 
tions in solid elements, J. Appl. Phys., 44, 
2470-2474, 1973. 

Marsh, S., ed., LASL Shock Hugoniot Data, 
University of Chicago Press, 1980. 

High-Pressure Phase 

Transformations in Selenium and 


H. K. Mao, G. Zou, and P. M. Bell 

Determination of the behavior of the 
elements at high pressure is an essen- 
tial part of any study of the earth's 
core. Elemental selenium (Se) and tellu- 
rium (Te), both group VI(B) elements, 
have unusual high-pressure forms that 
follow a trend in their electrical proper- 

* Research supported in part by National Sci- 
ence Foundation grant EAR8008804 and De- 
partment of Energy grant DE-AS05-80ER10754. 

ties with increasing pressure from in- 
sulator, to semiconductor, to metallic. 
Furthermore, the metallic forms that 
are quenchable to 1 bar become super- 
conducting at cryogenic temperatures, 
and thus their properties have been 
studied extensively in recent years 
(Bundy and Dunn, 1979). Most studies 
involved measurements of electrical re- 
sistance; however, some of the actual 
phases have not been characterized, 
and their stability fields have been de- 
termined only approximately. The find- 
ing of a new phase of Te and three new 
phases of Se, all determined by in situ 
x-ray diffraction analysis at high pres- 
sure, is reported herein. 

Powders of high purity (99.999 w